SB 1154: California Preventing Algorithmic Collusion Act of 2024.
- Session Year: 2023-2024
- House: Senate
- Latest Version Date: 2024-03-18
Current Status:
Failed
(2024-04-17: Set for hearing April 23.)
Introduced
In Committee
First Chamber
In Committee
Second Chamber
Enacted
Existing law establishes the Attorney General as the head of the Department of Justice, with charge of all legal matters in which the state is interested, except as specified. Existing law imposes various requirements on the Attorney General related to consumer protection, including, among others, the supervision of charitable trusts and the enforcement of antitrust laws. Existing law, commonly known as the Cartwright Act, identifies certain acts that are unlawful restraints of trade and unlawful trusts and prescribes provisions for its enforcement through civil actions.
This bill, the California Preventing Algorithmic Collusion Act of 2024 (the act), would require a person, as defined, upon request of the Attorney General, to provide to the Attorney General a written report, as specified, on each pricing algorithm, as defined, identified in the request. The act would require that all information submitted in a report under these provisions be treated as confidential and be considered privileged and confidential trade secrets exempt from disclosure under the California Public Records Act. The act would authorize the Attorney General to share the report with the National Institute of Standards and Technology for technical assistance in understanding the report, as specified.
This act would prohibit a person from using or distributing any pricing algorithm that uses, incorporates, or was trained with nonpublic competitor data, as defined. If the Attorney General has reason to believe that a person has violated these provisions, the act would authorize the Attorney General to bring a civil action against the person in any court of competent jurisdiction in this state, as specified. The act would provide that certain presumptions are established as applicable, including that the defendant entered into a contract in restraint of trade, as specified, if the Attorney General establishes any of specified requirements, including that the defendant distributed the pricing algorithm to 2 or more persons with the intent that the pricing algorithm be used to set or recommend a price, as defined, or commercial term, as defined, of a product or service in the same market or a related market and 2 or more persons used the pricing algorithm to set or recommend a price or commercial term of a product or service.
This act would require a person that has $5,000,000 or more in annual revenue that uses a pricing algorithm to recommend or set a price or commercial term to make certain disclosures, as specified, including to a customer, before the customer purchases the relevant product or service, that the price or a commercial term is set or recommended by a pricing algorithm. The act would provide that failure to comply with these disclosure provisions constitutes an unfair trade practice, as specified. If the Attorney General has reason to believe that a person has violated these disclosure provisions, the act authorizes the Attorney General to bring a civil action against the person in any court of competent jurisdiction in this state, as specified.
ExistingExisting law establishes the Governors Office of Business and Economic Development (GO-Biz), which serves the Governor as the lead entity for economic strategy and the marketing of California on issues relating to business development, private sector investment, and economic growth. Existing law establishes the Department of Financial Protection and Innovation in the Business, Consumer Services, and Housing Agency, headed by the Commissioner of Financial Protection and Innovation. Under existing law, the department has charge of the execution of specified laws relating to various financial institutions and financial services. Existing law requires the commissioner to publish on the departments internet website an annual report detailing actions taken during the prior year, including, among others, recommendations intended to result in improved oversight, greater transparency, or increased availability of beneficial financial products and services in the marketplace.
This bill would, commencing January 1, 2026, and every two years thereafter, require the Attorney General, in collaboration with the Department of Financial Protection and Innovation, to submit a report to the Legislature on the progress of emerging markets in California, as specified. The bill would make the report requirement inoperative on January 1, 2030, and would make related findings and declarations.This act would require, on or before January 1, 2027, the office of the Attorney General to publish on its internet website, and notify the Legislature of the publication of, the results of a study conducted in collaboration with GO-Biz and the Department of Financial Protection and Innovation on the use of pricing algorithms using, incorporating, or trained with either or both public and nonpublic data, as specified, including, among other information, the prevalence of pricing algorithms.
Existing constitutional provisions require that a statute that limits the right of access to the meetings of public bodies or the writings of public officials and agencies be adopted with findings demonstrating the interest protected by the limitation and the need for protecting that interest.
This bill would make legislative findings to that effect.
Existing law establishes the Attorney General as the head of the Department of Justice, with charge of all legal matters in which the state is interested, except as specified. Existing law imposes various requirements on the Attorney General related to consumer protection, including, among others, the supervision of charitable trusts and the enforcement of antitrust laws.
Existing law establishes the Department of Financial Protection and Innovation in the Business, Consumer Services, and Housing Agency, headed by the Commissioner of Financial Protection and Innovation. Under existing law, the department has charge of the execution of specified laws relating to various financial institutions and financial services. Existing law requires the commissioner to publish on the departments internet website an annual report detailing actions taken during the prior year, including, among others, recommendations intended to result in improved oversight, greater transparency, or increased availability of beneficial financial products and services in the marketplace.
This bill would, commencing January 1, 2026, and every two years thereafter, require the Attorney General, in collaboration with the Department of Financial Protection and Innovation, to submit a report to the Legislature on the progress of emerging markets in California, as specified. The bill would make the report requirement inoperative on January 1, 2030, and would make related findings and declarations.
This chapter shall be known, and may be cited, as the California Preventing Algorithmic Collusion Act of 2024.
17371.For purposes of this chapter, the following definitions apply:
(a)Antitrust laws has the same meaning as defined in the Clayton Act (15 U.S.C. Sec. 12), and includes Section 45 of Title 15 of the United States Code, and this part, including provisions commonly known as the Cartwright Act (Chapter 2 (commencing with Section 16700)).
(b)Commercial term means any of the following:
(1)Level of service.
(2)Availability.
(3)Output, including quantities of products produced or distributed or the amount or level of service provided.
(4)Rebates or discounts made available.
(c)Distribute, distribution, and distributing include selling, licensing, providing access to, or otherwise making available by any means, including through a subscription or the sale of a service.
(d)Nonpublic competitor data means nonpublic data that is derived from or otherwise provided by another person that competes in the same market as a person, or a related market. Nonpublic competitor data does not include information distributed, reported, or otherwise communicated in a way that does not reveal any underlying data from a competitor, such as narrative industry reports, news reports, business commentaries, or generalized industry survey results.
(e)Nonpublic data means information that is not widely available or easily accessible to the public, including information about prices, commercial terms, and related products or services, regardless of whether the data is attributable to a specific competitor or anonymized.
(f)Person has the same meaning as defined in Section 16702.
(g)Price means the amount of money or other thing of value, whether tangible or not, expected, required, or given in payment for any product or service, including compensation paid to an employee or independent contractor for services provided.
(h)Pricing algorithm means any computational process, including a computational process derived from machine learning or other artificial intelligence techniques, that processes data to recommend or set a price or commercial term within the jurisdiction of this state.
17372.(a)Upon written request by the Attorney General, a person shall, no later than 30 days after the date of the written request or any later date approved by the Attorney General, provide to the Attorney General a written report on each pricing algorithm identified in the request.
(b)Each report pursuant to subdivision (a) shall include all of the following:
(1)Information on whether the person is responsible for the development or distribution of the pricing algorithm, or whether a third party is responsible for the development or distribution of the pricing algorithm, including the identity and contact information of any other person responsible for the development or distribution of the pricing algorithm.
(2)Information on whether the pricing algorithm autonomously sets prices or commercial terms and whether there is human review of any recommendation or decision of the pricing algorithm.
(3)An explanation of the rules or processes that the pricing algorithm uses to set or recommend prices or commercial terms.
(4)A description of all data the pricing algorithm uses to set or recommend prices or commercial terms, including data used to train the algorithm.
(5)All sources and collection processes, including the frequency of collection, of any data that the pricing algorithm uses to set or recommend prices or commercial terms.
(6)Whether the pricing algorithm engages in price discrimination by setting or recommending different prices or commercial terms for the following:
(A)Different customers seeking identical or nearly identical products or services, and if so, the factors used in differentiating among those customers.
(B)Different employees or independent contractors providing substantially similar services, and if so, the factors used in differentiating among those employees or independent contractors.
(7)Any changes made to the pricing algorithm between the date of receipt of the request under subdivision (a) and the date of certification under subdivision (c).
(c)The chief executive officer, chief economist, chief technology officer, or a corporate officer of similar authority of a person shall certify, under penalty of perjury, the accuracy of a report under subdivision (a) submitted by the person.
(d)All information submitted in a report under subdivision (a) shall be treated as confidential and shall be considered to be privileged and confidential trade secrets exempt from disclosure under the California Public Records Act (Division 10 (commencing with Section 7920.000) of Title 1 of the Government Code).
(e)Notwithstanding subdivision (d), the Attorney General may share the report pursuant to subdivision (a) with the National Institute of Standards and Technology for technical assistance in understanding the report if the National Institute of Standards and Technology agrees to refrain from disclosing the contents of a report shared under this subdivision or the analysis of the report by the National Institute of Standards and Technology to any person, except the office of the Attorney General.
(f)This section shall not be construed to do either of the following:
(1)Limit the ability of the Attorney General to issue a civil investigative demand, to issue a subpoena, to seek discovery in the course of litigation, or to otherwise obtain information through other means available to the Attorney General.
(2)Restrict the use of information submitted in a report under subdivision (a) in the course of a formal investigation, enforcement action, litigation, trial, or other proceeding, in accordance with the confidentiality procedures applicable to that proceeding.
17373.(a)A person shall not use or distribute any pricing algorithm that uses, incorporates, or was trained with nonpublic competitor data.
(b)If the Attorney General has reason to believe that a person has violated subdivision (a), the Attorney General may bring a civil action against the person in any court of competent jurisdiction in this state to seek to recover one or both of the following:
(1)A civil penalty of one of the following:
(A)Not less than ten thousand dollars ($10,000), adjusted for inflation on the basis of the California Consumer Price Index, for each day during which the violation occurs or continues to occur.
(B)The sum of the price of each product or service sold using the pricing algorithm in violation of subdivision (a).
(2)Other appropriate relief, including an injunction or other equitable relief.
17374.(a)If the Attorney General meets the requirements of subdivision (b), there shall be, as applicable, the following presumptions:
(1)The defendant entered into a contract in restraint of trade in violation of Chapter 1 (commencing with Section 16600).
(2)The defendant entered into a contract, agreement, or trust in violation of Chapter 2 (commencing with Section 16700).
(3)The defendant entered into a contract in violation of Chapter 4 (commencing with Section 17040).
(b)The applicable presumptions described in subdivision (a) shall apply if the Attorney General establishes either of the following:
(1)The defendant distributed the pricing algorithm to two or more persons with the intent that the pricing algorithm be used to set or recommend a price or commercial term of a product or service in the same market or a related market and two or more persons used the pricing algorithm to set or recommend a price or commercial term of a product or service.
(2)The defendant used the pricing algorithm to set or recommend a price or commercial term of a product or service and the pricing algorithm was used by another person to set or recommend a price or commercial term of a product or service in the same market or a related market.
(c)The presumptions under subdivision (a) shall not apply to a defendant if the defendant did not develop or distribute the pricing algorithm and the defendant demonstrates by clear and convincing evidence that the defendant did not have actual knowledge or could not have reasonably known that the pricing algorithm used nonpublic competitor data.
(d)In a civil case in which a presumption described in subdivision (a) is applicable, any person that distributed the pricing algorithm and knew, or could have reasonably known, that the pricing algorithm would use, incorporate, or be trained with nonpublic competitor data shall be jointly and severally liable for any violation of Chapter 1, 2, or 4.
(e)Nothing in this section shall impair or limit the applicability of antitrust laws.
17375.(a)A person that has five million dollars ($5,000,000) or more in annual revenue that uses a pricing algorithm to recommend or set a price or commercial term shall make, in a clear manner, the following disclosures:
(1)To a customer, before the customer purchases the relevant product or service, that the price or a commercial term, as applicable, is set or recommended by a pricing algorithm.
(2)To a current or prospective employee or independent contractor that the price or commercial term for services rendered as an employee or independent contractor is set or recommended by a pricing algorithm.
(b)(1)If applicable, a disclosure under subdivision (a) shall state that the pricing algorithm sets or recommends different prices for the following:
(A)Different customers seeking identical or nearly identical products or services.
(B)Employees or independent contractors providing substantially similar services.
(2)If applicable, a disclosure under subdivision (a) shall do both of the following:
(A)State that the pricing algorithm was developed or distributed by a person other than the person making the disclosure.
(B)Provide the identity of the person that developed or distributed the pricing algorithm.
(c)Failure to provide a disclosure under subdivision (a), including the information required under subdivision (b), shall constitute an unfair trade practice in violation of Chapter 4.
(d)If the Attorney General has reason to believe that a person has violated subdivision (a) or (b), the Attorney General may bring a civil action against the person in any court of competent jurisdiction in this state to seek to recover one or both of the following:
(1)A civil penalty of not less than five thousand dollars ($5,000), adjusted for inflation on the basis of the California Consumer Price Index, for each day during which the violation occurs or continues to occur.
(2)Other appropriate relief, including an injunction or other equitable relief.
(e)Nothing in this section shall impair or limit the applicability of antitrust laws.
17376.On or before January 1, 2027, notwithstanding Section 10231.5 of the Government Code, the office of the Attorney General shall publish on its internet website, and notify the Legislature of the publication of, the results of a study conducted in collaboration with the Governors Office of Business and Economic Development and the Department of Financial Protection and Innovation on the use of pricing algorithms using, incorporating, or trained with either or both public and nonpublic data, including the following information:
(a)The prevalence of pricing algorithms.
(b)The frequency of the use of pricing algorithms to engage in price or wage discrimination.
(c)The potential for persons to use pricing algorithms to engage in behavior that increases prices, lowers wages, reduces output, lowers quality, deters innovation, or otherwise harms the competitive process outside of the price-fixing context.
(d)The potential benefits or efficiencies of pricing algorithms.
(e)Any industries, sectors, or markets in which pricing algorithms may warrant additional oversight or regulation to protect competition and consumers.
(f)Recommendations for additional legislation, regulation, or rulemaking relating to competition and consumer protection issues arising from the use of pricing algorithms.
SEC. 2.The Legislature finds and declares that Section 1 of this act, which adds Chapter 8 (commencing with Section 17370) to Part 2 of Division 7 of the Business and Professions Code, imposes a limitation on the publics right of access to the meetings of public bodies or the writings of public officials and agencies within the meaning of Section 3 of Article I of the California Constitution. Pursuant to that constitutional provision, the Legislature makes the following findings to demonstrate the interest protected by this limitation and the need for protecting that interest:
To protect the privileged and confidential trade secrets and commercial or financial information of persons submitting a written report, it is necessary to restrict the publics access to this information.
SECTION 1.The Legislature finds and declares:(a)Californias economy is dynamic and diverse, with emerging markets playing a significant role in contributing to the states economic growth and development.(b)The state recognizes the importance of safeguarding consumers in emerging markets where novel products and services may pose unique risks. A thorough understanding of these markets is essential to establish robust consumer protection measures.(c)Emerging markets often witness the introduction of new brands and trademarks. Studying these markets is imperative to maintain the integrity of trademark systems, prevent infringement, and ensure fair competition among businesses.(d)California serves as an innovation hub, fostering the growth of emerging technologies and industries. Studying emerging markets within the state is essential to stay abreast of technological advancements and protect intellectual property rights.(e)Emerging markets provide unique opportunities for small businesses and startups. Understanding these markets is essential for creating a supportive environment that encourages entrepreneurship and sustainable growth.(f)The fluid nature of emerging markets demands continuous study to comprehend market dynamics, identify potential risks, and establish effective risk management strategies to protect businesses and consumers alike.(g)Recognizing the critical importance of data security, California acknowledges the necessity to protect consumer data from unauthorized use, theft, and exploitation by entities. Emerging markets may pose increased risks of data vulnerabilities, requiring proactive measures to safeguard sensitive information.(h)California is committed to preventing scenarios where consumer data is stolen and sold overseas, potentially allowing other countries to gain insights into the states needs and vulnerabilities. That exploitation undermines Californias sovereignty and necessitates stringent measures to prevent unauthorized access and data transfer.(i)California aims to counteract situations where data stolen from the state is sold back by other countries, enabling them to influence or manipulate local markets. Stringent regulations and international cooperation are essential to prevent such practices and maintain control over the states data.SEC. 2.Section 12535 is added to the Government Code, immediately following Section 12534, to read:12535.(a)Commencing January 1, 2026, and every two years thereafter, the Attorney General, in collaboration with the Department of Financial Protection and Innovation, shall submit a report to the Legislature on the progress of emerging markets in California. The report shall include information on all of the following:(1)Data governance and security, including an assessment of data governance practices within emerging markets, addressing issues related to data security and preventing unauthorized use or exploitation of consumer data.(2)Anticompetitive practices, including an examination of potential anticompetitive practices within emerging markets with consideration given to the impact of data sharing and roll-up strategies on competition, output, and pricing.(3)Consumer protection measures, including an overview of measures taken to protect consumers within emerging markets, including safeguards against fraud, misinformation, and other risks associated with novel products and services.(4)Small business support and innovation, including an evaluation of initiatives supporting the growth of small businesses and startups in emerging markets and an assessment of the impact of data practices on fostering innovation.(5)Equity considerations, including an analysis of policies and programs designed to promote equity within emerging markets, ensuring that economic benefits are distributed equitably across diverse communities and demographic groups.(6)Federal collaboration, including an overview of collaboration with federal partners to establish global data protection standards applicable to emerging markets.(7)Roll-up strategies oversight, including an examination of roll-up strategies, particularly those involving private equity firms or strategic buyers, with a focus on monitoring and enforcing potentially anticompetitive roll-ups. Consideration of prior notice and approval requirements for future transactions in consent orders.(b)(1)The requirement for submitting a report imposed under subdivision (a) is inoperative on January 1, 2030, pursuant to Section 10231.5 of the Government Code.(2)A report to be submitted pursuant to subdivision (a) shall be submitted in compliance with Section 9795 of the Government Code.The Legislature finds and declares:
(a)Californias economy is dynamic and diverse, with emerging markets playing a significant role in contributing to the states economic growth and development.
(b)The state recognizes the importance of safeguarding consumers in emerging markets where novel products and services may pose unique risks. A thorough understanding of these markets is essential to establish robust consumer protection measures.
(c)Emerging markets often witness the introduction of new brands and trademarks. Studying these markets is imperative to maintain the integrity of trademark systems, prevent infringement, and ensure fair competition among businesses.
(d)California serves as an innovation hub, fostering the growth of emerging technologies and industries. Studying emerging markets within the state is essential to stay abreast of technological advancements and protect intellectual property rights.
(e)Emerging markets provide unique opportunities for small businesses and startups. Understanding these markets is essential for creating a supportive environment that encourages entrepreneurship and sustainable growth.
(f)The fluid nature of emerging markets demands continuous study to comprehend market dynamics, identify potential risks, and establish effective risk management strategies to protect businesses and consumers alike.
(g)Recognizing the critical importance of data security, California acknowledges the necessity to protect consumer data from unauthorized use, theft, and exploitation by entities. Emerging markets may pose increased risks of data vulnerabilities, requiring proactive measures to safeguard sensitive information.
(h)California is committed to preventing scenarios where consumer data is stolen and sold overseas, potentially allowing other countries to gain insights into the states needs and vulnerabilities. That exploitation undermines Californias sovereignty and necessitates stringent measures to prevent unauthorized access and data transfer.
(i)California aims to counteract situations where data stolen from the state is sold back by other countries, enabling them to influence or manipulate local markets. Stringent regulations and international cooperation are essential to prevent such practices and maintain control over the states data.
SEC. 2.Section 12535 is added to the Government Code, immediately following Section 12534, to read:12535.(a)Commencing January 1, 2026, and every two years thereafter, the Attorney General, in collaboration with the Department of Financial Protection and Innovation, shall submit a report to the Legislature on the progress of emerging markets in California. The report shall include information on all of the following:
(1)Data governance and security, including an assessment of data governance practices within emerging markets, addressing issues related to data security and preventing unauthorized use or exploitation of consumer data.
(2)Anticompetitive practices, including an examination of potential anticompetitive practices within emerging markets with consideration given to the impact of data sharing and roll-up strategies on competition, output, and pricing.
(3)Consumer protection measures, including an overview of measures taken to protect consumers within emerging markets, including safeguards against fraud, misinformation, and other risks associated with novel products and services.
(4)Small business support and innovation, including an evaluation of initiatives supporting the growth of small businesses and startups in emerging markets and an assessment of the impact of data practices on fostering innovation.
(5)Equity considerations, including an analysis of policies and programs designed to promote equity within emerging markets, ensuring that economic benefits are distributed equitably across diverse communities and demographic groups.
(6)Federal collaboration, including an overview of collaboration with federal partners to establish global data protection standards applicable to emerging markets.
(7)Roll-up strategies oversight, including an examination of roll-up strategies, particularly those involving private equity firms or strategic buyers, with a focus on monitoring and enforcing potentially anticompetitive roll-ups. Consideration of prior notice and approval requirements for future transactions in consent orders.
(b)(1)The requirement for submitting a report imposed under subdivision (a) is inoperative on January 1, 2030, pursuant to Section 10231.5 of the Government Code.
(2)A report to be submitted pursuant to subdivision (a) shall be submitted in compliance with Section 9795 of the Government Code.
|
Date
|
Status
|
|---|---|
| Apr 17, 2024 | Set for hearing April 23. |
| Apr 16, 2024 | April 16 set for first hearing canceled at the request of author. |
| Apr 2, 2024 | Set for hearing April 16. |
| Mar 18, 2024 | From committee with author's amendments. Read second time and amended. Re-referred to Com. on JUD. |
| Feb 21, 2024 | Referred to Com. on JUD. |
| Feb 15, 2024 | From printer. May be acted upon on or after March 16. |
| Feb 14, 2024 | Introduced. Read first time. To Com. on RLS. for assignment. To print. |
This list shows positions taken on this bill by organizations on the date displayed, which is linked to the original source.
Positions by individuals are not included.
|
Organization
|
Position
|
Date
|
|---|---|---|
| California Chamber Of Commerce | * Against | Apr 23, 2024 |
| California Credit Union League | ** Against | |
| CALIFORNIA FUELS & CONVENIENCE ALLIANCE | ** Against | |
| California Retailers Association | ** Against | |
| Chamber Of Progress | ** Against | |
| Children's Advocacy Institute | ** Support | Apr 23, 2024 |
| Computer And Communications Industry Association | * Against | Apr 23, 2024 |
| Insights Association | ** Against | |
| Los Angeles Area Chamber Of Commerce | ** Against | |
| NetChoice | ** Against | |
| Oakland Privacy | ** Support | |
| Organization for Social Media Safety | ** Support | Apr 23, 2024 |
| Software & Information Industry Association | ** Against | |
| Technet | ** Against | Apr 23, 2024 |
| The Travel Technology Association | ** Against | |
| Victims of Illicit Drugs | ** Support | Apr 23, 2024 |
Senator Thomas Umberg, Chair
2023-2024 Regular Session
Bill No: SB 1154 (Author:Hurtado)
Version: March 18, 2024
Hearing Date: April 23, 2024
Fiscal: Yes
Urgency: No
Consultant: AWM
SUBJECT
California Preventing Algorithmic Collusion Act of 2024
DIGEST
This bill establishes the California Preventing Algorithmic Collusion Act of 2024, which prohibits the use of pricing algorithms to set or recommend a price or commercial term in this state that incorporates nonpublic competitor data, as defined, which may be enforced by the Attorney General; establishes a partially rebuttable presumption that the use of a prohibited pricing algorithm is a violation of specified state laws prohibiting anticompetitive behavior; and adds additional provisions relating to the Attorney General’s investigation of, and disclosures of the use of, pricing algorithms.
EXECUTIVE SUMMARY
The federal Sherman Act and the state’s Cartwright Act impose broad prohibitions on price-fixing arrangements between competitors. While competitors may factor into their pricing decisions publicly available information about their competitors’ prices or plans, the law prohibits attempts to circumvent the competitive forces of the market by colluding with competitors to keep prices high or supply low.
Recently, some developers have devised pricing algorithms that collect nonpublic data from competitors and make recommendations on pricing or other terms based on that information. While it may seem obvious that using an algorithm to combine nonpublic competitor data and set prices is not meaningfully different, as a legal matter, than having the competitors get together in person and trade pricing information, some of these algorithms were allowed to proliferate for years. More recently, however, the United States Department of Justice (USDOJ) and the Federal Trade Commission (FTC) have stepped up their enforcement of federal antitrust laws in this area, including filing a lawsuit against an agricultural commodities price-setting service and filing a statement of information in a pending suit against a rent-setting service. California has joined as a party in the pending suit, and representatives of the California Department of Justice have announced their intention to ramp up California’s Cartwright Act enforcement measures.
Against this backdrop, this bill purports to create a wholly new legal mechanism prohibiting pricing algorithms that incorporate nonpublic competitor data and permitting the Attorney General to investigate and enforce violations. The bill also requires all persons or entities with $5 million or more in revenue to disclose every time a price or term was set using a pricing algorithm, and requires the California Department of Justice, in consultation with the Governor’s Office of Business and Economic Development and the Department of Financial Protection and Innovation (DFPI), to submit a report on the use of pricing algorithms.
This bill is sponsored by the author and is supported by Oakland Privacy. This bill is opposed by the California Chamber of Commerce, the California Credit Union League, the California Fuels & Convenience Alliance, the California Retailers Association, the Chamber of Progress, the Computer & Communications Industry Association, the Insights Association, the Los Angeles Area Chamber of Commerce, NetChoice, the Software & Information Industry Association, TechNet, and The Travel Technology Association.
PROPOSED CHANGES TO THE LAW
Existing federal law:
- Establishes the Sherman Antitrust Act of 1890 (Sherman Act). (15 U.S.C. §§ 1-7.)
- Makes illegal, under the Sherman Act, every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the states or with foreign nations. (15 U.S.C. § 1.)
- Authorizes a state attorney general to bring a civil action in the name of the state in any district court of the United States having jurisdiction over the defendant to secure monetary relief, as provided, for violations of the Sherman Act. (15 U.S.C. § 15c.)
- Establishes the Clayton Act. (15 U.S.C. §§ 12-27.)
- Defines, under the Clayton Act, “antitrust laws” to include the Sherman Act, certain provisions of the Wilson Tariff Act, and the Clayton Act, as amended. (15 U.S.C. § 12.)
- Makes illegal, under the Clayton Act, certain exclusive dealing agreements, tying contracts, corporate mergers and acquisitions, and interlocking directorates, as specified. (15 U.S.C. §§ 13-14.)
Existing state law:
- Prohibits that every contract by which anyone is restrained from engaging in a lawful profession, trade, or business of any kind is, to that extent, void, including specified noncompete clauses, subject to specified exemptions. (Bus. & Prof. Code, pt. 2, ch. 1, §§ 16600 et seq.)
- Establishes the Cartwright Act. (Bus. & Prof. Code, pt. 2, ch. 2, §§ 16700 et seq.)
- Defines a “trust” under the Cartwright Act as a combination of capital, skill, or acts by two or more persons for any of the following purposes:
- To create or carry out restrictions in trade or commerce.
- To limit or reduce the production, or increase the price of, merchandise or of any commodity.
- To prevent competition in manufacturing, making, transportation, sale, or purchase of merchandise, produce, or any commodity.
- To fix at any standard or figure, whereby its price to the public or consumer shall be in any manner controlled or established, any article or commodity of merchandise, produce, or commerce intended for sale, barter, use, or consumption in the state.
- To make or enter into or execute or carry out any contracts, obligations, or agreements of any kind or description, by which they do all or any combination of the following:
- Bind themselves not to sell, dispose of, or transport any article or any commodity or any article of trade, use, merchandise, commerce, or consumption below a common standard figure, or fixed value.
- Agree in any manner to keep the price of such article, commodity, or transportation at a fixed or graduated figure.
- Establish or settle the price of any article, commodity, or transportation between them or themselves and others, so as directly or indirectly to preclude a free and unrestricted competition among themselves, or any purchasers or consumers in the sale or transportation of any such article or commodity.
- Agree to pool, combine, or directly or indirectly unite any interests that they may have connected with the sale or transportation of any such article or commodity, that its price in any manner might be affected. (Bus. & Prof. Code, § 16720.)
- Makes every trust unlawful, against public policy, and void, except as exempted under the Cartwright Act. (Bus. & Prof. Code, § 16726.)
- Establishes the Unfair Practices Act (UPA), which is intended to safeguard the public against the creation or perpetuation of monopolies and to foster and encourage competition, by prohibiting unfair, dishonest, deceptive, destructive, fraudulent and discriminatory practices by which fair and honest competition is destroyed or prevented. (Bus. & Prof. Code, pt. 2, ch. 4, §§ 17000 et seq.)
- Prohibits, under the UPA, a range of behavior that reduces competition in pricing, including specified locality discrimination in pricing, sales under costs or loss leaders made with the intent of injuring competitors or destroying competition, and contracts for the performance of warranty service and repair below the cost of the service or repair. (Bus. & Prof. Code, §§ 17040-17051.)
- Permits the head of any department in the state to make investigations and prosecute actions concerning violations of law; as part of those investigations, the department head may inspect and copy books, records, and other items, issue subpoenas for the attendance of witnesses and the production of documents or other tangible things, and promulgate interrogatories. (Gov. Code, §§ 11180, 11181.)
- For purposes of investigating potential violations of the Cartwright Act and the UPA, extends all of the investigative powers granted to the Attorney General pursuant to 7) to the district attorney of any county when the district attorney reasonably believes that a violation has occurred. (Bus. & Prof. Code, § 16759(a).)
This bill:
- Establishes the California Preventing Algorithmic Collusion Act of 2024 within the Business and Professions Code.
- Defines the following terms:
- “Antitrust laws” has the same meaning as in the Clayton Act.
- “Commercial term” means any of the following: level of service; availability; output, including quantities of products produced or distributed or the amount or level of service provided; or rebates or discounts made available.
- “Distribute,” “distribution,” and “distributing” include selling, licensing, providing access to, or otherwise making available by any means, including through a subscription or the sale of a service
- “Nonpublic competitor data” means nonpublic data that is derived from or otherwise provided by another person that competes in the same market as a person, or a related market. “Nonpublic competitor data” does not include information distributed, reported, or otherwise communicated in a way that does not reveal any underlying data from a competitor, such as narrative industry reports, news reports, business commentaries, or generalized industry survey results.
- “Nonpublic data” means information that is not widely available or easily accessible to the public, including information about process, commercial terms, and related products or services, regardless of whether the data is attributable to a specific competitor or anonymized.
- “Person” has the same meaning as in Business and Professions Code section 16702: corporations, firms, partnerships, and associations existing under or authorized by the laws of this State or any other State, or any foreign country.
- “Price” means the amount of money or other thing of value, whether tangible or not, expected, required, or given in payment for any product or service, including compensation paid to an employee or independent contractor for services provided.
- “Pricing algorithm” means any computational process, including a computational process derived from machine learning or other artificial intelligence techniques, that process data to recommend or set a price or commercial term within the jurisdiction of this state.
- Requires a person, upon written request by the Attorney General, no later than 30 days after the date of the written request or any later date approved by the Attorney General, to provide to the Attorney General a written report on each pricing algorithm identified in the request.
- Requires a report made pursuant to 3) to include all of the following:
- Information on whether the person is responsible for the development or distribution of the pricing algorithm, or whether a third party is responsible for the development or distribution of the pricing algorithm, including the identity and contact information of any other person responsible for the development or distribution of the pricing algorithm.
- Information on whether the pricing algorithm autonomously sets prices or commercial terms and whether there is human review of any recommendation or decision of the pricing algorithm.
- An explanation of the rules or processes that the pricing algorithm uses to set or recommend prices or commercial terms.
- A description of all data the pricing algorithm uses to set or recommend prices or commercial terms, including data used to train the algorithm.
- All sources and collection processes, including the frequency of collection, of any data that the pricing algorithm uses to set or recommend prices or commercial terms.
- Whether the pricing algorithm engages in price discrimination by setting or recommending different prices or commercial terms for the following:
- Different customers seeking identical or nearly identical products or services, and if so, the factors used in differentiating among those customers.
- Different employers or independent contractors providing substantially similar services, and if so, the factors used in differentiating among those employees or independent contractors.
- Any changes made to the pricing algorithm between the date of receipt of the request under 3) and the date of certification under 5).
- Requires the chief executive officer, chief economist, chief technology officer, or a corporate officer of similar authority to certify, under penalty of perjury, the accuracy of the information submitted under 3).
- Provides that all information submitted in a report under 3) shall be treated as confidential and shall be considered to be privileged and confidential trade secrets exempt from disclosure under the California Public Records Act (CPRA); however, the Attorney General may share the report with the National Institute of Standards and Technology (NIST) for technical assistance in understanding the report if the NIST agrees to refrain from disclosing the contents of the report or the NIST’s analysis of the report to any person except the office of the Attorney General.
- Provides that 3)-7) shall not be construed to do either of the following:
- Limit the Attorney General to issue a civil investigative demand, to issue a subpoena, to seek discovery in the course of litigation, or to otherwise obtain information through other means available to the Attorney General.
- Restrict the use of information submitted in a report submitted under 3) in the course of a formal investigation, enforcement action, litigation, trial, or other proceeding, in accordance with the confidentiality procedures applicable to that proceeding.
- Prohibits a person from using or distributing any pricing algorithm that uses, incorporates, or was trained with nonpublic competitor data.
- Provides that, if the Attorney General has reason to believe that a person has violated 8), the Attorney General may bring a civil action against the person in any court of competent jurisdiction in the state to seek to recover one or both of the following:
- A civil penalty of either (1) not less than $10,000, adjusted for inflation on the basis of the California Consumer Price Index, for each day during which the violation occurs or continues to occur; or (2) the sum of the price of each product or service sold using the pricing algorithm in violation of 8).
- Other appropriate relief, including an injunction or other equitable relief.
- Provides that, if the Attorney General satisfies 11), there shall be, as applicable, the following presumptions:
- The defendant entered into a contract in restraint of trade in violation of Business and Professions Code sections 16600 et seq.
- The defendant entered into a contract, agreement or trust in violation of the Cartwright Act.
- The defendant entered into a contract in violation of the UPA.
- Provides that the presumptions in 10) shall apply if the Attorney General establishes either of the following:
- The defendant distributed the pricing algorithm to two or more persons with the intent that the pricing algorithm be used to set or recommend a price or commercial terms of a product or service in the same market or a related market and two or more persons used the pricing algorithm to set or recommend a price or commercial term of a product or service.
- The defendant used the pricing algorithm to set or recommend a price or commercial term of a product or service and the pricing algorithm was used by another person to set or recommend a price or commercial term of a product or service in the same market or related market.
- Provides that the presumptions in 10) shall not apply to a defendant if the defendant did not develop or distribute the pricing algorithm and the defendant demonstrates by clear and convincing evidence that the defendant did not have actual knowledge or could not have reasonably known that the pricing algorithm used nonpublic competitor data.
- Provides that, in a civil case in which a presumption in 10) is applicable, any person that distributed the pricing algorithm and knew, or could have reasonably known, that the pricing algorithm would use, incorporate, or be trained with nonpublic competitor data shall be jointly and severally liable for any violation of any violation of the statutes set forth in 10)(a)-(c).
- Provides that nothing in 10)-13) shall impair or limit the applicability of antitrust laws.
- Requires a person that has $5,000,000 or more in annual revenue that uses a pricing algorithm to recommend or set a price or commercial term to make, in a clear manner, the following disclosures:
- To a customer, before the customer purchases the relevant product or service, that the price or a commercial term, as applicable, is set or recommended by a pricing algorithm.
- To a current or prospective employee or independent contractor that the price or commercial term for services rendered as an employee or independent contractor is set or recommended by a pricing algorithm.
- If applicable, a statement that the pricing algorithm sets or recommends different prices for the following:
- Different customers seeking identical or nearly identical products or services.
- Employees or independent contractors providing substantially similar services.
- If applicable, a statement that the pricing algorithm was developed or distributed by a person other than the person making the disclosure, and the identity of the person that developed or distributed the pricing algorithm.
- Provides that the failure to provide a disclosure under 15) or any required components of the disclosure shall constitute a violation of the UPA.
- Permits the Attorney General, if they have any reason to believe that a person has violated 15), to bring a civil action against the person in any court of competent jurisdiction in the state to recover one or both of the following:
- A civil penalty of not less than $5000, adjusted for inflation on the basis of the California Consumer Price Index, for each day during which the violation occurs or continues to occur.
- Other appropriate relief, including an injunction or other equitable relief.
- Provides that nothing in 15)-17) shall impair or limit the applicability of antitrust laws.
- Provides that, on or before January 1, 2027, the Office of the Attorney General shall publish on its website, and notify the Legislature of the publication, the results of a study conducted with the Governor’s Office of Business and Economic Development and the DFPI on the use of pricing algorithms using, incorporating, or trained with either both public and nonpublic data, including the following information:
- The prevalence of pricing algorithms.
- The frequency and the use of pricing algorithms to engage in price or wage discrimination.
- The potential for persons to use pricing algorithms to engage in behavior that increases prices, lowers wages, reduces output, lowers quality, deters innovation, or otherwise harms the competitive process outside of the price-fixing context.
- The potential benefits or efficiencies of pricing algorithms.
- Any industries, sectors, or markets in which pricing algorithms may warrant additional oversight or regulation to protect competition and consumers.
- Recommendations for additional legislation, regulation, or rulemaking relating to competition and consumer protection issues arising from the use of pricing algorithms.
- Finds and declares that the exemption to the CPRA for information disclosed to the Attorney General under 6) is needed to protect the privileged and confidential trade secrets and commercial or financial information of persons submitting a written report.
COMMENTS
- Author’s comment
In today's digital age, pricing algorithms wield significant power in determining product costs. However, they also open the door to collusion and price-fixing among businesses, threatening fair competition and consumer welfare.
SB 1154, the Preventing Algorithmic Collusion Act, tackles this issue head-on by requiring businesses to disclose details about their pricing algorithms. By shedding light on development, data sources, and discriminatory practices, the bill ensures transparency and accountability in pricing decisions.
Moreover, SB 1154 bans the use of pricing algorithms fueled by nonpublic competitor data, leveling the playing field and deterring anti-competitive behavior. It also mandates disclosure for businesses with substantial revenue, empowering consumers with crucial information.
Additionally, the bill commissions a study by the Attorney General to assess the impact of pricing algorithms, providing valuable insights for future policymaking.
In essence, SB 1154 safeguards fair competition and consumer interests in the digital marketplace. It promotes transparency, accountability, and innovation, essential for a thriving economy. I urge support for this crucial legislation to protect Californians and uphold market integrity.
2. Antitrust law and algorithms
Under the federal Sherman Act,[1] “[horizontal] price-fixing agreements are unlawful per se.”[2] This per se bar extends to any “combination formed for the purpose and with the effect of raising, depressing, fixing, pegging, or stabilizing the price of a commodity” in interstate or foreign commerce.[3] Likewise, under the State’s own antitrust law, the Cartwright Act,[4] “agreements fixing or tampering with prices are illegal per se.”[5] These prohibitions “rest on the premise that unrestrained interaction of competitive forces will yield the best allocation of our economic resources, the lowest prices, the highest quality and the greatest material progress, while at the same time providing an environment conducive to the preservation of our democratic political and social institutions.”[6] The Cartwright Act’s prohibitions on anticompetitive behavior are “broader and deeper in range” than the federal Sherman Act’s.[7]
“Pricing algorithms are intended to help firms determine optimal pricing on a near real-time basis.”[8] A human setting prices has to (1) take in new information, (2) analyze the effect of the new information on their own prices, and (3) determine whether to raise or lower prices, and by how much. A pricing algorithm, on the other hand, uses artificial intelligence and machine learning “to weigh variables such as supply and demand, competitor pricing, and delivery time,” as well as any other factors its programmers have baked into the formula, and can set new prices nearly instantaneously in response to new information.[9] In recent years, developers have devised pricing algorithms that do more than incorporate public information, and instead collect nonpublic information from competitors in the market in order to make pricing recommendations.
Former FTC Chairwoman Maureen K. Ohlhausen has opined that the use of a vendor that provides algorithmic pricing services derived from confidential pricing information provided by multiple competitors is not, as an antitrust matter, new territory.[10] Instead, she opined, such services are merely an updated version of a long-prohibited practice, known as the “hub-and-spoke conspiracy”: “[j]ust as the antitrust laws do not allow competitors to exchange competitively sensitive information directly in an effort to stabilize or control industry pricing, they also prohibit using an intermediary to facilitate the exchange of confidential business information.”[11] To understand why this is such an easy call, Ohlhausen recommended replacing “algorithm” with “ ‘a guy named Bob’ ”:
Is it ok for a guy named Bob to collect confidential price strategy information from all the participants in a market, and then tell everybody how they should price? If it isn’t ok for a guy named Bob to do it, then it probably isn’t ok for an algorithm to do it either.[12]
This approach reflects the longstanding antitrust principle that “competitors cannot simply get around antitrust liability by acting through a third-party intermediary or joint venture.”[13]
In March of this year, the FTC and the USDOJ filed a Statement of Interest in a pending antitrust class action suit against RealPage, a company that developed an algorithm to recommend rents and occupancy levels for property owners.[14] According to the USDOJ, the allegations in the complaint—that RealPage used private pricing data from competitors to recommend prices to other competitors—constitute “per se unlawful conduct” on the part of the landlords using RealPage data:
Although not every use of an algorithm to set price qualifies as a per se violation of Section 1 of the Sherman Act, it is per se unlawful when, as alleged here, competitors knowingly combine their sensitive, nonpublic pricing and supply information in an algorithm that they rely upon in making pricing decisions, with the knowledge and expectation that other competitors will do the same.[15]
The USDOJ has also filed a civil antitrust lawsuit against Agri Stats, Inc., alleging that the business used nonpublic competitor data to recommend prices to participating meat producers.[16] California and several other attorneys general joined the Agri Stats lawsuit in November of last year.[17] This year, Senior Assistant Attorney General Paula Blizzard announced that the California Office of the Attorney General Antitrust Section is reviving its criminal antitrust program under the Cartwright Act.[18]
3. This bill creates an entirely new legal regime regarding the use of pricing algorithms that incorporate nonpublic competitor data
The bill is based on a pending piece of federal legislation, the Preventing Algorithmic Collusion Act of 2024, which was introduced this year and is sitting in the first committee.[19] The bills are virtually identical except for differences in the enforcing actor (e.g., the California Attorney General vs. the FTC) and the bill’s limitation to pricing algorithms used within the State.
Broadly speaking, the bill prohibits the use of any pricing algorithms that “uses, incorporates, or was trained on” nonpublic competitor data; only the Attorney General is empowered to enforce this prohibition, which must be done through a civil action filed in the superior court. The bill also provides that the use of prohibited algorithms establishes a partially rebuttable presumption that the use is a violation of the state’s prohibition on noncompetitive contracts, the Cartwright Act, and the UPA. Additionally, the bill establishes a new investigation mechanism for the Attorney General to obtain information about the potential use of prohibited algorithms and requires the Department of Justice to publish a study, in collaboration with the Governor’s Office of Business and Economic Development and the DFPI, on the use of pricing algorithms that use, incorporate, or are trained on public and/or nonpublic data. Finally, the bill requires a person or entity that has $5,000,000 more in annual revenue that uses a pricing algorithm to recommend or set a price or commercial term to disclose that use to consumers or employees or potential employees.
This bill, and the author’s comments about it, seem to run contrary to the legal consensus that the use of nonpublic competitor data in pricing algorithms is already prohibited under state law, including the Cartwright Act and the UPA. As discussed above, the USDOJ and Attorney General Bonta have already taken action against entities alleged to be providing collusive pricing algorithms. It is thus not clear why this bill’s prohibition on algorithms using nonpublic competitor data is necessary. Stakeholders have expressed concern that this bill could lend credence to an argument that collusion via algorithm is not already covered by existing law; this Committee’s understanding is that existing law does, in fact, prohibit competitors from combining nonpublic information in an algorithm to set prices.
Similarly, the Attorney General already has broad enforcement powers to conduct investigations, so it is unclear why the bill’s new investigatory mechanism is needed.[20]
There are also concerns about the scope of the bill. Many of the bill’s opponents note that the bill’s definitions are broad and vague, giving rise to the likelihood that the bill will sweep in pricing algorithms that are not the product of competitor collusion. For example, the bill defines “nonpublic data” as “information that is not widely available or easily accessible to the public, including information about process, commercial terms, and related products or services, regardless of whether the data is attributable to a specific competitor or anonymized.” This gives rise to many questions—how easily accessible does information have to be to be public? Can the information be behind a paywall? If the data are anonymized, how does a developer know if the information is widely available or not?—and provides little guidance to developers and sellers on how they can use pricing algorithms that comply with this law. Likewise, the bill prohibits the use of algorithms using, incorporating, or trained on nonpublic data from actors in a “related market,” but it is unclear how close two markets have to be in order to be deemed “related.”
The bill’s disclosure requirement also gives rise to concerns about overwarning. Warnings that are too prolific or warn about minor dangers run the risk of causing the public to “become bored or cynical and cease to pay attention all together.”[21] The benefits of any consumer warning, therefore, must be balanced against “ ‘the dangers of overwarning and of less meaningful warnings crowding out necessary warnings.’ ”[22] This bill mandates that every person or entity with $5 million more disclose to a consumer every time that a pricing algorithm was used to set the price. Given that pricing algorithms are used by major online retailers and in a wide array of markets,[23] it is unclear what benefit consumers would gain from this disclosure.
Finally, overall, there is a concern that drafting a bill relating to a fast-moving technology like pricing algorithms will ultimately hamper the Attorney General’s ability to take action against collusion. The Cartwright Act is deliberately broad so as to capture all possible forms of anticompetitive behavior.[24] This bill, on the other hand, could provide a road map to bad actors for how to evade enforcement, which would hamstring the Attorney General until the Legislature got around to passing another bill. It is unclear why this bill is preferable to the straightforward simplicity of the Cartwright Act. Moreover, as discussed above, the Office of the Attorney General has announced its intention to step up its criminal antitrust enforcement; this bill could interfere with, and maybe even harm, ongoing investigations.
5. Arguments in support
According to Oakland Privacy:
We support the California Legislature strengthening consumer protections and modernizing antitrust legislation. We hope that SB 1154 will help foster true innovation and competition, protect consumers and their privacy, and bolster antitrust compliance enforcement. Businesses are leveraging technology to make pricing decisions such as benchmarking, determining price elasticity and optimization based on a variety of variables including competitor behaviors, costs, and other market factors. It’s important to note that while these technologies are touted as a means of ensuring fair and competitive pricing in the marketplace, the ultimate goal of business is to maximize revenue and capture market share and therefore pricing technology will be utilized as a means to those ends. Moreover while pricing technologies themselves and their applications are opaque, there is already evidence of collusion like price fixing and self preferencing. In experiments, pricing algorithms have already demonstrated collusion tactics without explicit programming to do so. This also highlights the need to factor in transparency and accountability not only with specific properties of algorithms, but also self-learning AI.
Leveraging AI for pricing utilizes large quantities of consumer and other data including behaviors and characteristics that are extremely privacy invasive. As evident by the overwhelming public support for the creation of the California Privacy Protection Agency, consumers are not in favor of having their data harvested. Specifically with pricing, according to a 2023 study, only 34% of consumers think restaurant dynamic pricing is good for customers, 52% think of it as price gouging, and 63% said it made it harder to budget spending.
6. Arguments in opposition
According to a coalition of ten of the bill’s opponents:
First and foremost, this bill is wholly unnecessary because price collusion is clearly already illegal under current federal and state laws. Existing antitrust laws prohibit competitors from colluding through common use of a third-party company to set prices by improperly using competitively sensitive information from rivals, and the prohibition applies regardless of the form the alleged collusion takes. In other words, whether it is by salespeople conspiring or computers running algorithms, collusion is collusion. Indeed, the very circumstances that appear to have inspired federal bills on this topic (which SB 1154 seems to track in large degree) have already been the subject of multiple litigations making their way through the courts and the U.S. Department of Justice’s Antitrust Division recently joined the effort[…]
SB 1154 additionally creates the presumption that a business has engaged in certain illicit activities, including that the business entered into a contract in restraint of trade, if it: (1) distributes a pricing algorithm to two or more persons with the intent that it be used to set or recommend a price or commercial term of a product or service in the same market or a related market, and two or more persons use the pricing algorithm to do so; or (2) uses a pricing algorithm to set or recommend a price or commercial term of a product or service and the pricing algorithm was used by another person to set or recommend a price or commercial term of a person in the same market or a related market (i.e., where there was no intent by the first business). To rebut this presumption, the business must not only show that it did not develop or distribute the pricing algorithm, but it must also demonstrate, by clear and convincing evidence, that the business neither knew, nor could it have reasonably known, that the pricing algorithm used “nonpublic competitor data.” Given that a business may not know until after the fact that the public information used was not widely known or easily accessible, rebutting these presumptions becomes incredibly challenging and highly unlikely.
Furthermore, if a business has annual revenues greater than $5 million and uses pricing algorithms to set a price or commercial term, SB 1154 requires that it also comply with certain disclosures that can be rather cumbersome to provide, particularly for smaller businesses, and that create liability exposure. While the disclosures must be made “in a clear manner” for a business to be considered compliant, nothing in the bill provides clarity as to what is and is not considered clear for purposes of compliance. Any failure to meet this requirement will automatically constitute an unfair trade practice subject to monetary penalties of $5,000 for each day the violation occurs or continues to occur and/or injunctive or “other equitable relief”. Meaning, it is possible that a business provides the disclosure in a manner that it believes is clear, but if the AG disagrees, the business will have violated the law. As a result of such confusion and liability exposure, businesses will likely restrict their use of pricing algorithms, which will undermine the benefits for businesses and consumers alike.
SUPPORT
Oakland Privacy
OPPOSITION
California Chamber of Commerce
California Credit Union League
California Fuels & Convenience Alliance
California Retailers Association
Chamber of Progress
Computer & Communications Industry Association
Insights Association
Los Angeles Area Chamber of Commerce
Netchoice
Software & Information Industry Association
TechNet
The Travel Technology Association
RELATED LEGISLATION
Pending Legislation:
SB 1047 (Wiener, 2024) requires developers of powerful artificial intelligence models and those providing the computing power to train such models to put appropriate safeguards and policies into place to prevent critical harms; and establishes a state entity to oversee the development of these models and calls for the creation of a public cloud computing cluster. This bill is pending before the Senate Governmental Organization Committee.
AB 2930 (Bauer-Kahan, 2024) prohibits a person or entity from using an automated decision tool, including an algorithm, in a way that results in algorithmic discrimination, as defined, in employment, educational, housing, and other contexts. AB 2930 is pending before the Assembly Privacy and Consumer Protection Committee.
Prior Legislation:
AB 331 (Bauer-Kahan, 2023) was largely similar to AB 2930 (Bauer-Kahan, 2024) and would have prohibited a person or entity from using an automated decision tool, including an algorithm, in a way that results in algorithmic discrimination, as defined, in employment, educational, housing, and other contexts. AB 331 died in the Assembly Appropriations Committee.
AB 2224 (McCarty, 2022) would have required online real estate platforms, known as iBuyers, that use algorithms to determine the value of a property and make offers to purchase a home without the use of a mortgage or other type of financing, to work with a local real estate broker when selling and completing a sale of real property in California. AB 2224 died in the Senate Business, Professions and Economic Development Committee.
AB 1651 (Kalra, 2022) would have established limitations on the use of data-driven technologies in the workplace by requiring employers to notify workers prior to data collection, initiating electronic monitoring, and deploying algorithms, and required the technology be used pursuant to a valid business practice and be job-related and that employers conduct impact assessments with worker input for algorithms. AB 1651 died in the Assembly Privacy and Consumer Protection Committee.
**************
15 U.S.C. §§ 1-7. ↑
U.S. v. Socony-Vacuum Oil Co. (1940 310 U.S. 150, 218. ↑
Id. at p. 223. ↑
Bus. & Prof. Code, pt. 2, ch. 2, §§ 16700 et seq. ↑
Oakland-Alameda County Builders’ Exchange v. F. P. Lathrop Constr. Co. (1971) 4 Cal.3d 354, 363. ↑
Marin County Bd. of Realtors, Inc. v. Palsson (1976) 16 Cal.3d 920, 935 (internal quotation marks omitted). ↑
In re Cipro Cases I & II (2015) 61 Cal.4th 116, 160 (internal quotation marks omitted). ↑
Bertini & Koenigsberg, The Pitfalls of Pricing Algorithms, Harvard Business Review (Sept.-Oct. 2021), available at https://hbr.org/2021/09/the-pitfalls-of-pricing-algorithms. All links in this analysis are current as of April 11, 2024. ↑
Ibid. ↑
Remarks of Maureen K. Ohlhausen, Acting Chairwoman, U.S. Federal Trade Commission, “Should We Fear The Things That Go Beep In the Night? Some Initial Thoughts on the Intersection of Antitrust Law and Algorithmic Pricing,” Remarks from the Concurrences Antitrust Financial Sector Conference (May 23, 2017), p. 10, available at https://www.ftc.gov/news-events/news/speeches/should-we-fear-things-go-beep-night-some-initial-thoughts-intersection-antitrust-law-algorithmic. ↑
Ibid. ↑
Ibid. ↑
Am. Needle, Inc. v. Nat’l Football League (2010) 560 U.S. 183, 202 (cleaned up). ↑
Statement of the Interest of the United States of America, filed in Duffy v. Yardi Systems, Inc., et al. (D.Minn.) Case No. 2:23-cv-01391-RSL, Dkt. No. 149 (Mar. 1, 2024). ↑
Memorandum of Law in Support of the Statement of Interest of the United States, filed in Duffy v. Yardi Systems, Inc., et al. (D.Minn.) Case No. 2:23-cv-01391-RSL, Dkt. No. 149-2 (Mar. 1, 2024), p. 15. ↑
See U.S. v. Agri Stats, Inc., Complaint, Case No. 0:23-cv-03009-JRT-JFD, Dkt. No. 1 (Sept. 28, 2023). ↑
See California Department of Justice, Press Release: Attorney General Bonta Joins Lawsuit Against Agri Stats for Facilitating Meat Processors’ Unlawful Increase of Chicken, Pork, and Turkey Prices (Nov. 6, 2023), https://oag.ca.gov/news/press-releases/attorney-general-bonta-joins-lawsuit-against-agri-stats-facilitating-meat; U.S. v. Agri Stats, Inc., Second Amended Complaint, Case No. 0:23-cv-03009, Dkt. No. 50 (Nov. 15, 2023) ↑
Lau, et al., California revives criminal enforcement of the Cartwright Act, The Daily Journal (Mar. 27, 2024), available at https://www.dailyjournal.com/articles/377823-california-revives-criminal-enforcement-of-the-cartwright-act. ↑
See Sen. No. 3686, 118th Cong., 2d Sess. (2024). ↑
See Gov. Code, §§ 11180, 11181. The district attorney of any county is also empowered to investigate potential Cartwright Act and UPA violations, meaning county actors already have broader investigative authority than that granted by the bill. (See Bus. & Proc. Code, § 16759(a).) ↑
Noah, The Imperative to Warn: Distentagling the “Right to Know” from the “Need to Know” About Consumer Product Hazards, 11 Yale J. on Reg. 293, 384 (Summer 1994). ↑
Dowhal v. SmithKline Beecham Consumer Health Care (2004) 32 Cal.4th 910, 932. ↑
E.g., Wang, et al., Algorithms, Artificial Intelligence and Simple Rule-Based Pricing (Apr. 24, 2023), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4144905; Brown & McKay, Are online prices higher because of pricing algorithms?, Brookings (Jul. 7, 2022), https://www.brookings.edu/articles/are-online-prices-higher-because-of-pricing-algorithms/. ↑
Cianci v. Superior Court (1985) 40 Cal.3d 903, 917-918. ↑
Senator Thomas Umberg, Chair
2023-2024 Regular Session
Bill No: SB 1154 (Author:Hurtado)
Version: March 18, 2024
Hearing Date: April 16, 2024
Fiscal: Yes
Urgency: No
Consultant: AWM
SUBJECT
California Preventing Algorithmic Collusion Act of 2024
DIGEST
This bill establishes the California Preventing Algorithmic Collusion Act of 2024, which prohibits the use of pricing algorithms to set or recommend a price or commercial term in this state that incorporates nonpublic competitor data, as defined, which may be enforced by the Attorney General; establishes a partially rebuttable presumption that the use of a prohibited pricing algorithm is a violation of specified state laws prohibiting anticompetitive behavior; and adds additional provisions relating to the Attorney General’s investigation of, and disclosures of the use of, pricing algorithms.
EXECUTIVE SUMMARY
The federal Sherman Act and the state’s Cartwright Act impose broad prohibitions on price-fixing arrangements between competitors. While competitors may factor into their pricing decisions publicly available information about their competitors’ prices or plans, the law prohibits attempts to circumvent the competitive forces of the market by colluding with competitors to keep prices high or supply low.
Recently, some developers have devised pricing algorithms that collect nonpublic data from competitors and make recommendations on pricing or other terms based on that information. While it may seem obvious that using an algorithm to combine nonpublic competitor data and set prices is not meaningfully different, as a legal matter, than having the competitors get together in person and trade pricing information, some of these algorithms were allowed to proliferate for years. More recently, however, the United States Department of Justice (USDOJ) and the Federal Trade Commission (FTC) have stepped up their enforcement of federal antitrust laws in this area, including filing a lawsuit against an agricultural commodities price-setting service and filing a statement of information in a pending suit against a rent-setting service. California has joined as a party in the pending suit, and representatives of the California Department of Justice have announced their intention to ramp up California’s Cartwright Act enforcement measures.
Against this backdrop, this bill purports to create a wholly new legal mechanism prohibiting pricing algorithms that incorporate nonpublic competitor data and permitting the Attorney General to investigate and enforce violations. The bill also requires all persons or entities with $5 million or more in revenue to disclose every time a price or term was set using a pricing algorithm, and requires the California Department of Justice, in consultation with the Governor’s Office of Business and Economic Development and the Department of Financial Protection and Innovation (DFPI), to submit a report on the use of pricing algorithms.
This bill is sponsored by the author and is supported by Oakland Privacy. This bill is opposed by the California Chamber of Commerce, the California Credit Union League, the California Fuels & Convenience Alliance, the California Retailers Association, the Chamber of Progress, the Computer & Communications Industry Association, the Insights Association, the Los Angeles Area Chamber of Commerce, NetChoice, the Software & Information Industry Association, TechNet, and The Travel Technology Association.
PROPOSED CHANGES TO THE LAW
Existing federal law:
- Establishes the Sherman Antitrust Act of 1890 (Sherman Act). (15 U.S.C. §§ 1-7.)
- Makes illegal, under the Sherman Act, every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the states or with foreign nations. (15 U.S.C. § 1.)
- Authorizes a state attorney general to bring a civil action in the name of the state in any district court of the United States having jurisdiction over the defendant to secure monetary relief, as provided, for violations of the Sherman Act. (15 U.S.C. § 15c.)
- Establishes the Clayton Act. (15 U.S.C. §§ 12-27.)
- Defines, under the Clayton Act, “antitrust laws” to include the Sherman Act, certain provisions of the Wilson Tariff Act, and the Clayton Act, as amended. (15 U.S.C. § 12.)
- Makes illegal, under the Clayton Act, certain exclusive dealing agreements, tying contracts, corporate mergers and acquisitions, and interlocking directorates, as specified. (15 U.S.C. §§ 13-14.)
Existing state law:
- Prohibits that every contract by which anyone is restrained from engaging in a lawful profession, trade, or business of any kind is, to that extent, void, including specified noncompete clauses, subject to specified exemptions. (Bus. & Prof. Code, pt. 2, ch. 1, §§ 16600 et seq.)
- Establishes the Cartwright Act. (Bus. & Prof. Code, pt. 2, ch. 2, §§ 16700 et seq.)
- Defines a “trust” under the Cartwright Act as a combination of capital, skill, or acts by two or more persons for any of the following purposes:
- To create or carry out restrictions in trade or commerce.
- To limit or reduce the production, or increase the price of, merchandise or of any commodity.
- To prevent competition in manufacturing, making, transportation, sale, or purchase of merchandise, produce, or any commodity.
- To fix at any standard or figure, whereby its price to the public or consumer shall be in any manner controlled or established, any article or commodity of merchandise, produce, or commerce intended for sale, barter, use, or consumption in the state.
- To make or enter into or execute or carry out any contracts, obligations, or agreements of any kind or description, by which they do all or any combination of the following:
- Bind themselves not to sell, dispose of, or transport any article or any commodity or any article of trade, use, merchandise, commerce, or consumption below a common standard figure, or fixed value.
- Agree in any manner to keep the price of such article, commodity, or transportation at a fixed or graduated figure.
- Establish or settle the price of any article, commodity, or transportation between them or themselves and others, so as directly or indirectly to preclude a free and unrestricted competition among themselves, or any purchasers or consumers in the sale or transportation of any such article or commodity.
- Agree to pool, combine, or directly or indirectly unite any interests that they may have connected with the sale or transportation of any such article or commodity, that its price in any manner might be affected. (Bus. & Prof. Code, § 16720.)
- Makes every trust unlawful, against public policy, and void, except as exempted under the Cartwright Act. (Bus. & Prof. Code, § 16726.)
- Establishes the Unfair Practices Act (UPA), which is intended to safeguard the public against the creation or perpetuation of monopolies and to foster and encourage competition, by prohibiting unfair, dishonest, deceptive, destructive, fraudulent and discriminatory practices by which fair and honest competition is destroyed or prevented. (Bus. & Prof. Code, pt. 2, ch. 4, §§ 17000 et seq.)
- Prohibits, under the UPA, a range of behavior that reduces competition in pricing, including specified locality discrimination in pricing, sales under costs or loss leaders made with the intent of injuring competitors or destroying competition, and contracts for the performance of warranty service and repair below the cost of the service or repair. (Bus. & Prof. Code, §§ 17040-17051.)
- Permits the head of any department in the state to make investigations and prosecute actions concerning violations of law; as part of those investigations, the department head may inspect and copy books, records, and other items, issue subpoenas for the attendance of witnesses and the production of documents or other tangible things, and promulgate interrogatories. (Gov. Code, §§ 11180, 11181.)
- For purposes of investigating potential violations of the Cartwright Act and the UPA, extends all of the investigative powers granted to the Attorney General pursuant to 7) to the district attorney of any county when the district attorney reasonably believes that a violation has occurred. (Bus. & Prof. Code, § 16759(a).)
This bill:
- Establishes the California Preventing Algorithmic Collusion Act of 2024 within the Business and Professions Code.
- Defines the following terms:
- “Antitrust laws” has the same meaning as in the Clayton Act.
- “Commercial term” means any of the following: level of service; availability; output, including quantities of products produced or distributed or the amount or level of service provided; or rebates or discounts made available.
- “Distribute,” “distribution,” and “distributing” include selling, licensing, providing access to, or otherwise making available by any means, including through a subscription or the sale of a service
- “Nonpublic competitor data” means nonpublic data that is derived from or otherwise provided by another person that competes in the same market as a person, or a related market. “Nonpublic competitor data” does not include information distributed, reported, or otherwise communicated in a way that does not reveal any underlying data from a competitor, such as narrative industry reports, news reports, business commentaries, or generalized industry survey results.
- “Nonpublic data” means information that is not widely available or easily accessible to the public, including information about process, commercial terms, and related products or services, regardless of whether the data is attributable to a specific competitor or anonymized.
- “Person” has the same meaning as in Business and Professions Code section 16702: corporations, firms, partnerships, and associations existing under or authorized by the laws of this State or any other State, or any foreign country.
- “Price” means the amount of money or other thing of value, whether tangible or not, expected, required, or given in payment for any product or service, including compensation paid to an employee or independent contractor for services provided.
- “Pricing algorithm” means any computational process, including a computational process derived from machine learning or other artificial intelligence techniques, that process data to recommend or set a price or commercial term within the jurisdiction of this state.
- Requires a person, upon written request by the Attorney General, no later than 30 days after the date of the written request or any later date approved by the Attorney General, to provide to the Attorney General a written report on each pricing algorithm identified in the request.
- Requires a report made pursuant to 3) to include all of the following:
- Information on whether the person is responsible for the development or distribution of the pricing algorithm, or whether a third party is responsible for the development or distribution of the pricing algorithm, including the identity and contact information of any other person responsible for the development or distribution of the pricing algorithm.
- Information on whether the pricing algorithm autonomously sets prices or commercial terms and whether there is human review of any recommendation or decision of the pricing algorithm.
- An explanation of the rules or processes that the pricing algorithm uses to set or recommend prices or commercial terms.
- A description of all data the pricing algorithm uses to set or recommend prices or commercial terms, including data used to train the algorithm.
- All sources and collection processes, including the frequency of collection, of any data that the pricing algorithm uses to set or recommend prices or commercial terms.
- Whether the pricing algorithm engages in price discrimination by setting or recommending different prices or commercial terms for the following:
- Different customers seeking identical or nearly identical products or services, and if so, the factors used in differentiating among those customers.
- Different employers or independent contractors providing substantially similar services, and if so, the factors used in differentiating among those employees or independent contractors.
- Any changes made to the pricing algorithm between the date of receipt of the request under 3) and the date of certification under 5).
- Requires the chief executive officer, chief economist, chief technology officer, or a corporate officer of similar authority to certify, under penalty of perjury, the accuracy of the information submitted under 3).
- Provides that all information submitted in a report under 3) shall be treated as confidential and shall be considered to be privileged and confidential trade secrets exempt from disclosure under the California Public Records Act (CPRA); however, the Attorney General may share the report with the National Institute of Standards and Technology (NIST) for technical assistance in understanding the report if the NIST agrees to refrain from disclosing the contents of the report or the NIST’s analysis of the report to any person except the office of the Attorney General.
- Provides that 3)-7) shall not be construed to do either of the following:
- Limit the Attorney General to issue a civil investigative demand, to issue a subpoena, to seek discovery in the course of litigation, or to otherwise obtain information through other means available to the Attorney General.
- Restrict the use of information submitted in a report submitted under 3) in the course of a formal investigation, enforcement action, litigation, trial, or other proceeding, in accordance with the confidentiality procedures applicable to that proceeding.
- Prohibits a person from using or distributing any pricing algorithm that uses, incorporates, or was trained with nonpublic competitor data.
- Provides that, if the Attorney General has reason to believe that a person has violated 8), the Attorney General may bring a civil action against the person in any court of competent jurisdiction in the state to seek to recover one or both of the following:
- A civil penalty of either (1) not less than $10,000, adjusted for inflation on the basis of the California Consumer Price Index, for each day during which the violation occurs or continues to occur; or (2) the sum of the price of each product or service sold using the pricing algorithm in violation of 8).
- Other appropriate relief, including an injunction or other equitable relief.
- Provides that, if the Attorney General satisfies 11), there shall be, as applicable, the following presumptions:
- The defendant entered into a contract in restraint of trade in violation of Business and Professions Code sections 16600 et seq.
- The defendant entered into a contract, agreement or trust in violation of the Cartwright Act.
- The defendant entered into a contract in violation of the UPA.
- Provides that the presumptions in 10) shall apply if the Attorney General establishes either of the following:
- The defendant distributed the pricing algorithm to two or more persons with the intent that the pricing algorithm be used to set or recommend a price or commercial terms of a product or service in the same market or a related market and two or more persons used the pricing algorithm to set or recommend a price or commercial term of a product or service.
- The defendant used the pricing algorithm to set or recommend a price or commercial term of a product or service and the pricing algorithm was used by another person to set or recommend a price or commercial term of a product or service in the same market or related market.
- Provides that the presumptions in 10) shall not apply to a defendant if the defendant did not develop or distribute the pricing algorithm and the defendant demonstrates by clear and convincing evidence that the defendant did not have actual knowledge or could not have reasonably known that the pricing algorithm used nonpublic competitor data.
- Provides that, in a civil case in which a presumption in 10) is applicable, any person that distributed the pricing algorithm and knew, or could have reasonably known, that the pricing algorithm would use, incorporate, or be trained with nonpublic competitor data shall be jointly and severally liable for any violation of any violation of the statutes set forth in 10)(a)-(c).
- Provides that nothing in 10)-13) shall impair or limit the applicability of antitrust laws.
- Requires a person that has $5,000,000 or more in annual revenue that uses a pricing algorithm to recommend or set a price or commercial term to make, in a clear manner, the following disclosures:
- To a customer, before the customer purchases the relevant product or service, that the price or a commercial term, as applicable, is set or recommended by a pricing algorithm.
- To a current or prospective employee or independent contractor that the price or commercial term for services rendered as an employee or independent contractor is set or recommended by a pricing algorithm.
- If applicable, a statement that the pricing algorithm sets or recommends different prices for the following:
- Different customers seeking identical or nearly identical products or services.
- Employees or independent contractors providing substantially similar services.
- If applicable, a statement that the pricing algorithm was developed or distributed by a person other than the person making the disclosure, and the identity of the person that developed or distributed the pricing algorithm.
- Provides that the failure to provide a disclosure under 15) or any required components of the disclosure shall constitute a violation of the UPA.
- Permits the Attorney General, if they have any reason to believe that a person has violated 15), to bring a civil action against the person in any court of competent jurisdiction in the state to recover one or both of the following:
- A civil penalty of not less than $5000, adjusted for inflation on the basis of the California Consumer Price Index, for each day during which the violation occurs or continues to occur.
- Other appropriate relief, including an injunction or other equitable relief.
- Provides that nothing in 15)-17) shall impair or limit the applicability of antitrust laws.
- Provides that, on or before January 1, 2027, the Office of the Attorney General shall publish on its website, and notify the Legislature of the publication, the results of a study conducted with the Governor’s Office of Business and Economic Development and the DFPI on the use of pricing algorithms using, incorporating, or trained with either both public and nonpublic data, including the following information:
- The prevalence of pricing algorithms.
- The frequency and the use of pricing algorithms to engage in price or wage discrimination.
- The potential for persons to use pricing algorithms to engage in behavior that increases prices, lowers wages, reduces output, lowers quality, deters innovation, or otherwise harms the competitive process outside of the price-fixing context.
- The potential benefits or efficiencies of pricing algorithms.
- Any industries, sectors, or markets in which pricing algorithms may warrant additional oversight or regulation to protect competition and consumers.
- Recommendations for additional legislation, regulation, or rulemaking relating to competition and consumer protection issues arising from the use of pricing algorithms.
- Finds and declares that the exemption to the CPRA for information disclosed to the Attorney General under 6) is needed to protect the privileged and confidential trade secrets and commercial or financial information of persons submitting a written report.
COMMENTS
- Author’s comment
In today's digital age, pricing algorithms wield significant power in determining product costs. However, they also open the door to collusion and price-fixing among businesses, threatening fair competition and consumer welfare.
SB 1154, the Preventing Algorithmic Collusion Act, tackles this issue head-on by requiring businesses to disclose details about their pricing algorithms. By shedding light on development, data sources, and discriminatory practices, the bill ensures transparency and accountability in pricing decisions.
Moreover, SB 1154 bans the use of pricing algorithms fueled by nonpublic competitor data, leveling the playing field and deterring anti-competitive behavior. It also mandates disclosure for businesses with substantial revenue, empowering consumers with crucial information.
Additionally, the bill commissions a study by the Attorney General to assess the impact of pricing algorithms, providing valuable insights for future policymaking.
In essence, SB 1154 safeguards fair competition and consumer interests in the digital marketplace. It promotes transparency, accountability, and innovation, essential for a thriving economy. I urge support for this crucial legislation to protect Californians and uphold market integrity.
2. Antitrust law and algorithms
Under the federal Sherman Act,[1] “[horizontal] price-fixing agreements are unlawful per se.”[2] This per se bar extends to any “combination formed for the purpose and with the effect of raising, depressing, fixing, pegging, or stabilizing the price of a commodity” in interstate or foreign commerce.[3] Likewise, under the State’s own antitrust law, the Cartwright Act,[4] “agreements fixing or tampering with prices are illegal per se.”[5] These prohibitions “rest on the premise that unrestrained interaction of competitive forces will yield the best allocation of our economic resources, the lowest prices, the highest quality and the greatest material progress, while at the same time providing an environment conducive to the preservation of our democratic political and social institutions.”[6] The Cartwright Act’s prohibitions on anticompetitive behavior are “broader and deeper in range” than the federal Sherman Act’s.[7]
“Pricing algorithms are intended to help firms determine optimal pricing on a near real-time basis.”[8] A human setting prices has to (1) take in new information, (2) analyze the effect of the new information on their own prices, and (3) determine whether to raise or lower prices, and by how much. A pricing algorithm, on the other hand, uses artificial intelligence and machine learning “to weigh variables such as supply and demand, competitor pricing, and delivery time,” as well as any other factors its programmers have baked into the formula, and can set new prices nearly instantaneously in response to new information.[9] In recent years, developers have devised pricing algorithms that do more than incorporate public information, and instead collect nonpublic information from competitors in the market in order to make pricing recommendations.
Former FTC Chairwoman Maureen K. Ohlhausen has opined that the use of a vendor that provides algorithmic pricing services derived from confidential pricing information provided by multiple competitors is not, as an antitrust matter, new territory.[10] Instead, she opined, such services are merely an updated version of a long-prohibited practice, known as the “hub-and-spoke conspiracy”: “[j]ust as the antitrust laws do not allow competitors to exchange competitively sensitive information directly in an effort to stabilize or control industry pricing, they also prohibit using an intermediary to facilitate the exchange of confidential business information.”[11] To understand why this is such an easy call, Ohlhausen recommended replacing “algorithm” with “ ‘a guy named Bob’ ”:
Is it ok for a guy named Bob to collect confidential price strategy information from all the participants in a market, and then tell everybody how they should price? If it isn’t ok for a guy named Bob to do it, then it probably isn’t ok for an algorithm to do it either.[12]
This approach reflects the longstanding antitrust principle that “competitors cannot simply get around antitrust liability by acting through a third-party intermediary or joint venture.”[13]
In March of this year, the FTC and the USDOJ filed a Statement of Interest in a pending antitrust class action suit against RealPage, a company that developed an algorithm to recommend rents and occupancy levels for property owners.[14] According to the USDOJ, the allegations in the complaint—that RealPage used private pricing data from competitors to recommend prices to other competitors—constitute “per se unlawful conduct” on the part of the landlords using RealPage data:
Although not every use of an algorithm to set price qualifies as a per se violation of Section 1 of the Sherman Act, it is per se unlawful when, as alleged here, competitors knowingly combine their sensitive, nonpublic pricing and supply information in an algorithm that they rely upon in making pricing decisions, with the knowledge and expectation that other competitors will do the same.[15]
The USDOJ has also filed a civil antitrust lawsuit against Agri Stats, Inc., alleging that the business used nonpublic competitor data to recommend prices to participating meat producers.[16] California and several other attorneys general joined the Agri Stats lawsuit in November of last year.[17] This year, Senior Assistant Attorney General Paula Blizzard announced that the California Office of the Attorney General Antitrust Section is reviving its criminal antitrust program under the Cartwright Act.[18]
3. This bill creates an entirely new legal regime regarding the use of pricing algorithms that incorporate nonpublic competitor data
The bill is based on a pending piece of federal legislation, the Preventing Algorithmic Collusion Act of 2024, which was introduced this year and is sitting in the first committee.[19] The bills are virtually identical except for differences in the enforcing actor (e.g., the California Attorney General vs. the FTC) and the bill’s limitation to pricing algorithms used within the State.
Broadly speaking, the bill prohibits the use of any pricing algorithms that “uses, incorporates, or was trained on” nonpublic competitor data; only the Attorney General is empowered to enforce this prohibition, which must be done through a civil action filed in the superior court. The bill also provides that the use of prohibited algorithms establishes a partially rebuttable presumption that the use is a violation of the state’s prohibition on noncompetitive contracts, the Cartwright Act, and the UPA. Additionally, the bill establishes a new investigation mechanism for the Attorney General to obtain information about the potential use of prohibited algorithms and requires the Department of Justice to publish a study, in collaboration with the Governor’s Office of Business and Economic Development and the DFPI, on the use of pricing algorithms that use, incorporate, or are trained on public and/or nonpublic data. Finally, the bill requires a person or entity that has $5,000,000 more in annual revenue that uses a pricing algorithm to recommend or set a price or commercial term to disclose that use to consumers or employees or potential employees.
This bill, and the author’s comments about it, seem to run contrary to the legal consensus that the use of nonpublic competitor data in pricing algorithms is already prohibited under state law, including the Cartwright Act and the UPA. As discussed above, the USDOJ and Attorney General Bonta have already taken action against entities alleged to be providing collusive pricing algorithms. It is thus not clear why this bill’s prohibition on algorithms using nonpublic competitor data is necessary. Stakeholders have expressed concern that this bill could lend credence to an argument that collusion via algorithm is not already covered by existing law; this Committee’s understanding is that existing law does, in fact, prohibit competitors from combining nonpublic information in an algorithm to set prices.
Similarly, the Attorney General already has broad enforcement powers to conduct investigations, so it is unclear why the bill’s new investigatory mechanism is needed.[20]
There are also concerns about the scope of the bill. Many of the bill’s opponents note that the bill’s definitions are broad and vague, giving rise to the likelihood that the bill will sweep in pricing algorithms that are not the product of competitor collusion. For example, the bill defines “nonpublic data” as “information that is not widely available or easily accessible to the public, including information about process, commercial terms, and related products or services, regardless of whether the data is attributable to a specific competitor or anonymized.” This gives rise to many questions—how easily accessible does information have to be to be public? Can the information be behind a paywall? If the data are anonymized, how does a developer know if the information is widely available or not?—and provides little guidance to developers and sellers on how they can use pricing algorithms that comply with this law. Likewise, the bill prohibits the use of algorithms using, incorporating, or trained on nonpublic data from actors in a “related market,” but it is unclear how close two markets have to be in order to be deemed “related.”
The bill’s disclosure requirement also gives rise to concerns about overwarning. Warnings that are too prolific or warn about minor dangers run the risk of causing the public to “become bored or cynical and cease to pay attention all together.”[21] The benefits of any consumer warning, therefore, must be balanced against “ ‘the dangers of overwarning and of less meaningful warnings crowding out necessary warnings.’ ”[22] This bill mandates that every person or entity with $5 million more disclose to a consumer every time that a pricing algorithm was used to set the price. Given that pricing algorithms are used by major online retailers and in a wide array of markets,[23] it is unclear what benefit consumers would gain from this disclosure.
Finally, overall, there is a concern that drafting a bill relating to a fast-moving technology like pricing algorithms will ultimately hamper the Attorney General’s ability to take action against collusion. The Cartwright Act is deliberately broad so as to capture all possible forms of anticompetitive behavior.[24] This bill, on the other hand, could provide a road map to bad actors for how to evade enforcement, which would hamstring the Attorney General until the Legislature got around to passing another bill. It is unclear why this bill is preferable to the straightforward simplicity of the Cartwright Act. Moreover, as discussed above, the Office of the Attorney General has announced its intention to step up its criminal antitrust enforcement; this bill could interfere with, and maybe even harm, ongoing investigations.
5. Arguments in support
According to Oakland Privacy:
We support the California Legislature strengthening consumer protections and modernizing antitrust legislation. We hope that SB 1154 will help foster true innovation and competition, protect consumers and their privacy, and bolster antitrust compliance enforcement. Businesses are leveraging technology to make pricing decisions such as benchmarking, determining price elasticity and optimization based on a variety of variables including competitor behaviors, costs, and other market factors. It’s important to note that while these technologies are touted as a means of ensuring fair and competitive pricing in the marketplace, the ultimate goal of business is to maximize revenue and capture market share and therefore pricing technology will be utilized as a means to those ends. Moreover while pricing technologies themselves and their applications are opaque, there is already evidence of collusion like price fixing and self preferencing. In experiments, pricing algorithms have already demonstrated collusion tactics without explicit programming to do so. This also highlights the need to factor in transparency and accountability not only with specific properties of algorithms, but also self-learning AI.
Leveraging AI for pricing utilizes large quantities of consumer and other data including behaviors and characteristics that are extremely privacy invasive. As evident by the overwhelming public support for the creation of the California Privacy Protection Agency, consumers are not in favor of having their data harvested. Specifically with pricing, according to a 2023 study, only 34% of consumers think restaurant dynamic pricing is good for customers, 52% think of it as price gouging, and 63% said it made it harder to budget spending.
6. Arguments in opposition
According to a coalition of ten of the bill’s opponents:
First and foremost, this bill is wholly unnecessary because price collusion is clearly already illegal under current federal and state laws. Existing antitrust laws prohibit competitors from colluding through common use of a third-party company to set prices by improperly using competitively sensitive information from rivals, and the prohibition applies regardless of the form the alleged collusion takes. In other words, whether it is by salespeople conspiring or computers running algorithms, collusion is collusion. Indeed, the very circumstances that appear to have inspired federal bills on this topic (which SB 1154 seems to track in large degree) have already been the subject of multiple litigations making their way through the courts and the U.S. Department of Justice’s Antitrust Division recently joined the effort[…]
SB 1154 additionally creates the presumption that a business has engaged in certain illicit activities, including that the business entered into a contract in restraint of trade, if it: (1) distributes a pricing algorithm to two or more persons with the intent that it be used to set or recommend a price or commercial term of a product or service in the same market or a related market, and two or more persons use the pricing algorithm to do so; or (2) uses a pricing algorithm to set or recommend a price or commercial term of a product or service and the pricing algorithm was used by another person to set or recommend a price or commercial term of a person in the same market or a related market (i.e., where there was no intent by the first business). To rebut this presumption, the business must not only show that it did not develop or distribute the pricing algorithm, but it must also demonstrate, by clear and convincing evidence, that the business neither knew, nor could it have reasonably known, that the pricing algorithm used “nonpublic competitor data.” Given that a business may not know until after the fact that the public information used was not widely known or easily accessible, rebutting these presumptions becomes incredibly challenging and highly unlikely.
Furthermore, if a business has annual revenues greater than $5 million and uses pricing algorithms to set a price or commercial term, SB 1154 requires that it also comply with certain disclosures that can be rather cumbersome to provide, particularly for smaller businesses, and that create liability exposure. While the disclosures must be made “in a clear manner” for a business to be considered compliant, nothing in the bill provides clarity as to what is and is not considered clear for purposes of compliance. Any failure to meet this requirement will automatically constitute an unfair trade practice subject to monetary penalties of $5,000 for each day the violation occurs or continues to occur and/or injunctive or “other equitable relief”. Meaning, it is possible that a business provides the disclosure in a manner that it believes is clear, but if the AG disagrees, the business will have violated the law. As a result of such confusion and liability exposure, businesses will likely restrict their use of pricing algorithms, which will undermine the benefits for businesses and consumers alike.
SUPPORT
Oakland Privacy
OPPOSITION
California Chamber of Commerce
California Credit Union League
California Fuels & Convenience Alliance
California Retailers Association
Chamber of Progress
Computer & Communications Industry Association
Insights Association
Los Angeles Area Chamber of Commerce
Netchoice
Software & Information Industry Association
TechNet
The Travel Technology Association
RELATED LEGISLATION
Pending Legislation:
SB 1047 (Wiener, 2024) requires developers of powerful artificial intelligence models and those providing the computing power to train such models to put appropriate safeguards and policies into place to prevent critical harms; and establishes a state entity to oversee the development of these models and calls for the creation of a public cloud computing cluster. This bill is pending before the Senate Governmental Organization Committee.
AB 2930 (Bauer-Kahan, 2024) prohibits a person or entity from using an automated decision tool, including an algorithm, in a way that results in algorithmic discrimination, as defined, in employment, educational, housing, and other contexts. AB 2930 is pending before the Assembly Privacy and Consumer Protection Committee.
Prior Legislation:
AB 331 (Bauer-Kahan, 2023) was largely similar to AB 2930 (Bauer-Kahan, 2024) and would have prohibited a person or entity from using an automated decision tool, including an algorithm, in a way that results in algorithmic discrimination, as defined, in employment, educational, housing, and other contexts. AB 331 died in the Assembly Appropriations Committee.
AB 2224 (McCarty, 2022) would have required online real estate platforms, known as iBuyers, that use algorithms to determine the value of a property and make offers to purchase a home without the use of a mortgage or other type of financing, to work with a local real estate broker when selling and completing a sale of real property in California. AB 2224 died in the Senate Business, Professions and Economic Development Committee.
AB 1651 (Kalra, 2022) would have established limitations on the use of data-driven technologies in the workplace by requiring employers to notify workers prior to data collection, initiating electronic monitoring, and deploying algorithms, and required the technology be used pursuant to a valid business practice and be job-related and that employers conduct impact assessments with worker input for algorithms. AB 1651 died in the Assembly Privacy and Consumer Protection Committee.
**************
15 U.S.C. §§ 1-7. ↑
U.S. v. Socony-Vacuum Oil Co. (1940 310 U.S. 150, 218. ↑
Id. at p. 223. ↑
Bus. & Prof. Code, pt. 2, ch. 2, §§ 16700 et seq. ↑
Oakland-Alameda County Builders’ Exchange v. F. P. Lathrop Constr. Co. (1971) 4 Cal.3d 354, 363. ↑
Marin County Bd. of Realtors, Inc. v. Palsson (1976) 16 Cal.3d 920, 935 (internal quotation marks omitted). ↑
In re Cipro Cases I & II (2015) 61 Cal.4th 116, 160 (internal quotation marks omitted). ↑
Bertini & Koenigsberg, The Pitfalls of Pricing Algorithms, Harvard Business Review (Sept.-Oct. 2021), available at https://hbr.org/2021/09/the-pitfalls-of-pricing-algorithms. All links in this analysis are current as of April 11, 2024. ↑
Ibid. ↑
Remarks of Maureen K. Ohlhausen, Acting Chairwoman, U.S. Federal Trade Commission, “Should We Fear The Things That Go Beep In the Night? Some Initial Thoughts on the Intersection of Antitrust Law and Algorithmic Pricing,” Remarks from the Concurrences Antitrust Financial Sector Conference (May 23, 2017), p. 10, available at https://www.ftc.gov/news-events/news/speeches/should-we-fear-things-go-beep-night-some-initial-thoughts-intersection-antitrust-law-algorithmic. ↑
Ibid. ↑
Ibid. ↑
Am. Needle, Inc. v. Nat’l Football League (2010) 560 U.S. 183, 202 (cleaned up). ↑
Statement of the Interest of the United States of America, filed in Duffy v. Yardi Systems, Inc., et al. (D.Minn.) Case No. 2:23-cv-01391-RSL, Dkt. No. 149 (Mar. 1, 2024). ↑
Memorandum of Law in Support of the Statement of Interest of the United States, filed in Duffy v. Yardi Systems, Inc., et al. (D.Minn.) Case No. 2:23-cv-01391-RSL, Dkt. No. 149-2 (Mar. 1, 2024), p. 15. ↑
See U.S. v. Agri Stats, Inc., Complaint, Case No. 0:23-cv-03009-JRT-JFD, Dkt. No. 1 (Sept. 28, 2023). ↑
See California Department of Justice, Press Release: Attorney General Bonta Joins Lawsuit Against Agri Stats for Facilitating Meat Processors’ Unlawful Increase of Chicken, Pork, and Turkey Prices (Nov. 6, 2023), https://oag.ca.gov/news/press-releases/attorney-general-bonta-joins-lawsuit-against-agri-stats-facilitating-meat; U.S. v. Agri Stats, Inc., Second Amended Complaint, Case No. 0:23-cv-03009, Dkt. No. 50 (Nov. 15, 2023) ↑
Lau, et al., California revives criminal enforcement of the Cartwright Act, The Daily Journal (Mar. 27, 2024), available at https://www.dailyjournal.com/articles/377823-california-revives-criminal-enforcement-of-the-cartwright-act. ↑
See Sen. No. 3686, 118th Cong., 2d Sess. (2024). ↑
See Gov. Code, §§ 11180, 11181. The district attorney of any county is also empowered to investigate potential Cartwright Act and UPA violations, meaning county actors already have broader investigative authority than that granted by the bill. (See Bus. & Proc. Code, § 16759(a).) ↑
Noah, The Imperative to Warn: Distentagling the “Right to Know” from the “Need to Know” About Consumer Product Hazards, 11 Yale J. on Reg. 293, 384 (Summer 1994). ↑
Dowhal v. SmithKline Beecham Consumer Health Care (2004) 32 Cal.4th 910, 932. ↑
E.g., Wang, et al., Algorithms, Artificial Intelligence and Simple Rule-Based Pricing (Apr. 24, 2023), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4144905; Brown & McKay, Are online prices higher because of pricing algorithms?, Brookings (Jul. 7, 2022), https://www.brookings.edu/articles/are-online-prices-higher-because-of-pricing-algorithms/. ↑
Cianci v. Superior Court (1985) 40 Cal.3d 903, 917-918. ↑
Bill Author
Bill Co-Author(s):