Senate Standing Committee on Privacy, Digital Technologies, and Consumer Protection

Alright. Good afternoon. The Senate committee on privacy, digital technologies, and consumer protections come welcome to order. Welcome everyone. Today, we have no consent calendar and no further items have been pulled from the agenda.

Since we don't yet have a quorum, we'll begin as a subcommittee. And so we'll begin. First, we do have an author. Thank you. Assemblymember Bonta is here and she's gonna first present AB 1979.

Welcome and proceed when you're

ready. Thank you so much chair and committee members. First, I'd like to start by accepting the committee's amendments and thank the chair and committee staff for working with me on this bill. AB 1979 addresses the proliferation of artificial intelligence into health care in three very important ways.

First, it requires health facilities to ensure that no clinical decisions are being made solely by an output from a clinical decision support system, and that a licensed health care professional maintains the ability to exercise professional judgment in reviewing and approving that output.

Second, it prohibits the use of automated decision systems in health care settings to guide or instruct an unlicensed individual to do any clinical function that would require a license. And third, it clarifies provisions of the confidentiality of medical information act to ensure that direct to consumer health care chat bots that seek to access individuals medical records protect those records as otherwise required by law.

This year, we've seen major developers in consumer facing AIs, offer access to new tools to encourage individuals to connect their medical records and ask questions, purportedly to help them prepare for doctor's visits or get basic information. But they start to operate more like doctors and health care providers than they do like chat bots. We've also seen the rapid deployment of AI and health care and pending upheaval in our health care system from HR 1, where access to a human provider will become even more scarce.

Through the even just the work requirements of HR 1 alone, the California Healthcare Foundation estimates that 1.1,000,000 medical enrollees will lose coverage in the next few years, many due simply to administrative barriers. We are also facing the loss of potentially billions of health care dollars due to changes in federal financing rules, which financially strains providers. These losses will increase pressures on providers and health facilities as fewer people come in with insurance and there are less resources to provide care.

My concern is that there will be a temptation to turn AI to fill the needs to turn to AI to fill the needs of patients at reduced cost and loss of quality care. Automated decision systems and clinical decision support systems have been present in health care for decades.

They are tools that use algorithms, machine learning, and medical data to assist professionals with diagnostics, treatment, planning, and patient monitoring. I appreciate that. And I think it is at the innovation and at the skill that we need to ensure that our healthcare, industry can make sure to preserve and protect everyone. However, it's increasingly clear to me that the use of AI developing clinical decision support systems is evolving faster than our laws and regulations can keep up.

Protecting patient safety, keeping our professional workforce engaged in their work, and ensuring the integrity of health care will require us to not exclusively rely on AI to do the things quicker and cheaper, not while we are also still trying to work out and understand the potential risks along with the benefits.

This bill does that by ensuring that medical records are protected by direct to consumer health chat bots and that licensed health care professionals retain their ability to exercise professional judgment when reviewing the output of clinical decision support systems. Here to testify today are Dolores Trujillo, who is an NICU nurse with CNA Board and a CNA Board member and former president of the board of registered nursing and a representative from tech equity.

Welcome to you both and you'll each have two minutes.

Thank you. Good afternoon, chair and members. I'm Dolores Trujillo, board member of the California Nurses Association, the proud sponsor of AB 1979. I've been a registered nurse over twenty five years. I'm also the immediate past president of the board of registered nursing.

As nurses, our profession is both an art and a science. We care for people at some of their most vulnerable moments in their lives. As bedside nurses, we don't just execute tasks. We utilize the nursing process. We assess patients using our expertise, our eyes, ears and touch to detect the subtle nuances and changes of the human condition.

A nurse's professional health care judgment simply cannot be automated by an algorithm. This bill keeps patient care in human hands by ensuring AI supports licensed professionals instead of replacing or controlling our clinical judgment. Today, AI systems are being used in areas of patient care that require clinical judgment including patient assessment, clinical decision making, patient education, and a

hand off communication. For example, when I hand off a patient at the end of my shift, I'm not just passing along a summary of the patient's chart. I am using my clinical judgment to decide what matters most for the next nurse to know about my patient. Like what's changed and what worries me. AI can't predict how an individual patient will respond to care.

But if a hospital relies on AI to generate or determine that hand off communication, then an algorithm makes these judgment calls on care instead of the nurse. To be clear, this bill does not ban supportive technology. Nurses use technology every day. It simply reaffirms patient privacy laws and preserves the clear boundary around the care that California law reserves to license human healthcare professionals. Thank you and I respectfully ask for your aye vote.

Thank you.

Thank you. Before we proceed to your testimony, there's a magic moment in legislative committees when a quorum arrives that we don't miss because they can be fleeting unicorns running through the committee. So if you don't mind, we're gonna pause for just a moment to establish a quorum committee assistant. Would you please take the role?

Senators Cabaldon? Here. Cabaldon here. Jones?

I'm here.

Jones here. Gonzales McNerney?

Here.

McNerney here. Ochoa Bogh, Padilla? Here. Padilla here. Reyes?

Here. Reyes here. Umberg. Wiener.

Alright. A quorum is present. Please proceed.

It's great to be part of that magic moment. Good afternoon, chair and members. My name is Sam Gordon, chief advocacy officer at Tech Equity, and I'm here to speak in support of AB 1979. I think it's becoming clear and clear to all of us that the way employers design and deploy technology shapes not only who has power and agency at work, but whose expertise is valued and who is valued and who is accountable when things go wrong.

AI systems, as we all know, are now being introduced across industries to automate decisions, streamline workflows, and cut costs, and health care is obviously not an exception.

We believe that AB 1979 is a really important set of guardrails because it makes clear that AI systems cannot be deployed in a way that shifts clinical authority away from licensed professionals and into things like software systems or vendor assumptions about what might happen in that hospital or employer workflows. And importantly, the bill doesn't reject innovation.

I think it does the thing that all of us have said we want, which is it puts people next to technology, and it allows it to inform a human with expertise and their own judgment and allows it to be supportive but not determinative. And that allows that same thing that patients want, where your doctor or your nurse or your healthcare provider is still in the room, still making the decision, but benefiting from the patterns and analysis that technology can do.

And I think that this is a win win for patients, for workers, and for everything that we see that shows up in poll after poll that the public wants healthcare where their doctors, their healthcare professionals, their nurses are still in the room.

And so for those reasons, we would just ask that patient care doesn't become another site where automation is used to de skill work, cut corners, and improve or the sort of downgrade the patient experience. And so we wanna keep the promise of technology pointed in the right direction. We believe that AB 1979 does that. And for these reasons, we respectfully ask for your eye vote.

Alright. Thank you. Does anyone else wish to provide a testimony in support of the legislation? We invite you to come forward and and get and say your name, your affiliation if you have one, and your position on the bill.

Mister chair, member Sarah Flock, California Federation of Labor Unions in strong support. Holly Khan with California Nurses Association in support.

Jacqueline Montano, ICU nurse twenty years in strong support.

Shane Gussman on behalf of the Teamsters and the engineers and scientists of California in support.

Lynn Warmerdam, 37 year RN. Most of them at Doctors Medical Center in Modesto and strong support.

Good afternoon, chair members. JP Hanna with the California Nurses Association sponsored the bill. Thank you.

Alright. So no one else is there anyone that wishes to provide testimony, the lead lead testimony opposition to AB 1979? Welcome and you will also have two minutes.

Thank you. Good afternoon. Mark Farouk on behalf of the California Hospital Association. I do wanna begin by appreciating the committee amendments for moving the bill in the right direction. However, we do remain opposed unless amended.

Our remaining issues are with section three of the bill. For the other sections, our members already comply with those requirements. California's hospital shared the author's core principle. A licensed professional not software must hold final authority over patient care. We support requiring a clinician in the loop on any decision informed by an AI tool.

We start from agreement because hospitals are already deploying these tools to protect patients with Clinicians exercising judgement on every output. Early warning systems flag deterioration hours before overt clinical signs prompting a nurse or physician to assess and act. The tool prioritizes it does not diagnose. AI assisted imaging triage flags suspected strokes and alerts the stroke team where minutes saved translates directly in the preserved brain function.

AI is improving cancer screening primarily by acting as a second reader that flags suspicious regions for radiologists and pathologists rather than replacing their judgment.

Large trials have shown that AI assisted reading can increase cancer detection rates while reducing radiologists' workload with a clinician retaining final diagnostic authority. California's hospitals and health systems believe strongly in the responsible deployment of AI that is built on testing, training, compliance review, and ongoing monitoring to catch drift or bias.

Our remaining issues in 1979 are not about the concept of maintaining professional judgment, but rather we remain concerned about the burden to ensure that a licensed professional is utilizing their judgment given the penalties associated with the bill. And also remaining prohibitions that we are concerned may disrupt medical education and training. For those reasons, we remain opposed unless amended.

Thank you.

Thank you. Does anyone else wish to register their opposition to AB 1979? So please come forward to the stand up mic and share with us your name, your affiliation if you have one, and your position on the bill.

Thank you, mister chair and senators. MJDS in respectful opposition to the bill.

Thank you, chair and members. Ryan Perini on behalf of ATA Action in respectful posing that submitted position. Thank you.

Good afternoon. George Soares of the California Medical Association respectfully opposed unless amended. I line my comments with my colleague from CHA. Thank you. Alexis

Rodriguez of the California Chamber of Commerce also in an opposed unless amended position. Thank you.

Annalee Augustine with the Civil Justice Association of California also respectfully oppose unless amended. Thank you.

Good afternoon, chair members. Robert Boykin with TechNet in respectful oppose unless amended position.

Thank you, mister chair. Giacchion, Slick and Jensen on behalf of the Advanced Medical Technology Association, also oppose unless amended. Thank the author and staff for the community engagement. Appreciate it.

Charing committee members, Gilbert Laurie here with Biocom also with an opposed unless amended position. Thanks.

Apologies. MJDS on behalf of Kaiser Permanente in respectful opposition.

Lawrence Gaiden with the California Dental Association with a respectfully oppose post submitted position. Thank you.

Alright. Thanks to everyone who's provided testimony. Let's turn that to the committee for any questions or comments. Senator McNerney.

Well, thank you, mister chairman. I think the author bring this forward. This is parallel to my SB 947, the no robo bosses. Basically, which has calls for a human being in the loop on decisions of firing and firing and disappointing people. And this is similar.

You wanna have a human being in the loop. My reading of the hospital's association objection was that it would be that you wouldn't be able to use ADES's at all. But that isn't what you're trying to accomplish. You want a human being to be involved in in the decision making. I think that's perfectly appropriate.

So I'm gonna be supporting this bill. Thank you.

Senator Gomez Reyes.

Thank you. We also saw a similar bill for mental health professionals also being in the loop. It seems that there's agreement. I heard hearing from the hospital association. There's agreement that there should be a clinician, a nurse making in the loop to make sure that whatever decision is made, there is a human being that's involved in the process.

And who better than CNA to tell us what they can do and how they would do it. The I think that the only thing that I would say is that with the objections being posed by the hospitals, it seems that that is a discussion that continues at this point. Is that right?

That is fair. Absolutely. And, you know, me, Senator, that I will continue to work with the opposition to make sure that we hit a middle ground. I think the the opposition's examples are actually examples that I think where there's a secondary reading, where we, with the privacy committee's amendments, are, moving more towards specificity in terms of the scope of the bill where, we still are able to maintain professional judgment and the ability for a human to remain in the loop around, decision making.

What this bill seeks to do is ensure that there isn't any sole output from a clinical decision making, AI tool automated, tool that would exclude that professional from being involved in the in in the decision making.

I think it's it's clear that AI is evolving much faster than we can legislate. And I think these steps are important. I think that as we move forward making sure here that that the the stakeholders that are involved are involved in-in the legislation. But the final part of this which is the most important part of your your bill is to make sure that there is a human being.

There is a clinician that's going to make be involved in the loop to to look at whatever it is that is being provided by AI to make that final decision.

Thank you. Thank you.

Further questions or comments? I'm also gonna support the bill today for the reasons that the author's outlined but also a couple of others. The this domain is an interesting one like healthcare as a domain for this for this ongoing debate that we're having about the use of AI and automated decision systems in various in various other domains.

It's an interesting one because I I'm one of the reasons why I'm I'm excited about the bill is that that this this is the area with one of the greatest levels of promise around AI and automated decision systems. And I don't mean that in a sense that it's like ripe for technological change although there are aspects of that.

But unlike many other sectors of the economy where we're really, really worried about the impacts on workers and on safety and other things from the technologies. Health care happens in hospitals and medical practices and dental practices happen to be an area with lots and lots of overlapping guardrails for to use a 2026 overused term. There's a lot of institutions in the mix that don't exist on a on the factory warehouse or on a on a construction site or whatever. Whether that's state licensure of the facility.

Individual licensure of of virtually all of the Clinicians and training and standards and norms that they have gone through.

Professional organizations, accreditation. There's so many things that are there to design to assure that before just collect unions, they're quite present in in in many of the healthcare space. So there so there are already a lot of overlapping, institutions that are there to assure that we can't get this wrong. And also consumers quote unquote in healthcare unlike many other domains are highly motivated to care about the outcomes and the outcomes are often pretty measurable.

Whether I've survived or not, whether that rash is gone, whether my heart works again, they're they're more measurable than they might be in other domains.

So I so part of the reason why I think this is important here is that this this is an area healthcare is an area where where we where we actually have a chance to get this right. Not just ban it outright and also not just embrace it as though it's the second coming without thinking through how it's going to work. But it's an area where the various stakeholders and consumers and others have an opportunity to engage and to test and to try.

And I appreciate the the the bill has been crafted in a way that that does that. It doesn't even it doesn't even insist that every single decision has to have a human making it.

Right? So it's not that a human must be in the loop but number one, that when a human is making their decision and using their professional judgment subject to their licensure that it's that that is not being overwritten. And and also laying out the the systems by which the whole system itself has to be under the control. And so, I-I appreciate the work that's been done on this, by the author of the amendments.

I I think go go go quite far in trying to, trying to to make this workable.

Of course, very interested and to continue to hear from, many of the opponents about the details of Section three.

But I but I think the authors are taking a measured approach here to try to make sure that the overall interest of policy can, in conjunction with accreditation, licensure, collective bargaining, and everything else, can help this sector, lead to better outcomes for workers, for the institutions themselves, and of course, at the end of the day, you know, and the beginning of the day for for patients and and clients. And so with that, if there's no further discussion, is there a motion on this bill?

Move the bill.

Oh, did you have a comment, Senator Ocho Bo?

Oh, I just I didn't know you were gonna give an opportunity to speak but I can I can I can quickly make a statement?

Sure. Please please do.

If you're allowing.

Yes. Please please do. Requesting

me. Sorry. It just came back. I didn't I didn't know where we were. I was at a different committee.

So curious. It's my understanding that the California law already prohibits the delegation of practicing medicine to AI. Because the licensed physician must be the actual care provider. Is that correct?

I think that there is, language in, in guidelines that indicate that that is the case. I think that there is confusion about, whether AI should be treated as a provider, whether it is a product. And I think that the, legislation and the the law that we move forward with helps to, clarify that.

So I I think we are in a a greenfield space when it comes to AI and the deployment of AI, and this legislation seeks to at least, prepare some ability for us to have clarity around that.

Okay. Thank you, ma'am.

Alright. Was was there a motion I heard? Senator Reyes? Okay. Alright then.

Senator Vermaelte, would you like to close?

Thank you, chair. 93% of Americans have a concern about AI and health care, and the majority say that AI makes them trust health care less. At the same time, the some of the platform providers of their AI systems indicate that in a given week, you can have up to 230,000 individual queries related to healthcare. So we are at the point where people don't necessarily trust it, but people are using it at an exponential rate.

And also, we have a situation where, you know, given 1,357 medical devices using AI right now and authorized by the FDA.

60 FDA authorized medical devices using AI were linked to about a 182 product recalls. So there's a lot of opportunity for growth. We know that our I've heard by their own admission that AI developers are developing and reiterating and iterating while also deploying, which creates a situation where we have, unstable, products out there that have, to the chair's point, life and death consequences for, the people of California.

So I think it's very important that we provide an opportunity for us to be able to tackle this issue and do provide some regulations so that we have something to be able to hold on to both in the area of, this administration of that within a clinical setting and, within, the AI chatbot world. So with that, I appreciate the the conversation that we've had on this and I respectfully request an aye vote.

Alright. Then we'll proceed to the vote on AB 1979 committee assistant. Would you please call the roll?

The motion is do passed as amended to health. Senators Cabaldon.

Aye.

Cabaldon, aye. Jones? Gonzales? Aye. Gonzales, aye.

McNerney? Aye. McNerney, aye. Ochoa Bogh, Padilla?

Padilla, aye. Reyes?

Aye.

Aye.

Reyes, aye. Umberg? Weiner? Aye. Weiner, aye.

Ochoa Bog, no.

Six to one on call.

Alright. The vote is six to one. We'll place that measure on call. Next, we'll turn to AB 2624 also by assembly member Bonta. Please present when you're ready.

Thank you, mister chair and members. AB 2624, the Safe at Work Act, expands California's Safe at Home program to include immigrant service providers, their employees, and volunteers. This program allows for participants to use a substitute address designated by the secretary of state keeping their home, work, and school addresses out of the public record while still allowing them to safely receive mail and legal documents through the state.

This gives them a critical layer of protection and privacy in an environment where their personal safety is increasingly at risk. We know.

Individuals who provide vital immigrant support services, including legal aid, humanitarian relief, case management, and advocacy are facing targeted harassment. You will hear that from, those who will testify shortly. This is not hypothetical. Advocates and workers are receiving death threats, being targeted at courthouses, and facing coordinated online doxing campaigns, even facing this vitriol at their homes. These threats have risen sharply in 2025 on our and are expected to continue due to the current political climate.

At the same time, personal information is increasingly easy to access. Data brokers collect and sell information from public records, and social media can allow individuals to piece together identifying details. This makes it easier for bad actors to threaten or harm those who are simply doing their jobs. Advocates in California report doxing of staff and volunteers at immigration legal aid organizations, coordinated death threats against service providers, anti immigrant vigilante activity directed at organizations by name and address. This is-

the context we are facing. So organizations serving LGBTQ plus and immigrant communities commonly hide their locations, staff information, and other details to keep their teams and their people they serve safe. Currently, California state law does not provide adequate protections for their sensitive data and information, leaving immigrant advocates and service providers vulnerable. And generally, privacy act privacy laws act after the harm has already occurred and were not designed to address the coordinated online politically motivated harassment and hurt that we are now seeing.

This bill protects sensitive personal data in a way that empowers people to do their jobs safely and confidently under the secretary of State's safe at home program before harm occurs.

Since its inception nearly thirty years ago, the safe at home program has protected thousands of victims of stalking and domestic violence, along with reproductive health care workers. The confidentiality protection program provides participants with a substitute address to shield their real address, requires the secretary of state to act as an intermediary for mail and legal services, forwarding documents within a sit short try time frame, keeps individual participants home, work, and school address confidential, and protects them from public disclosure.

No one should face harassment or threats at their own home or at work. This legislation helps prevent individuals with malicious intent from targeting service providers where they live or work, reducing the risk of escalation to violence, bodily harm, or worse. At its core, this is about safeguarding the privacy, dignity, and the safety of immigrant service workers and their families.

California has already recognized in the context of reproductive health care providers that when program participants are targeted because of their work, the state has a responsibility to provide proactive protections. Immigrant service providers now face similar threats, and they deserve those same safeguards under the safe at home program. Without these safeguards, immigrant service providers may feel unsafe doing their jobs, which directly impacts families who depend on them for essential support.

I would like to note that, there are currently 8,000 people in the safe at home program under the secretary of state. There has not been one indication of there being anything unconstitutional about the safe at home program, and this bill is similarly straightforward.

It allows us to be able to extend the safe at home program to those people who are serving our very vulnerable communities and who find themselves in unsafe conditions. Here to testify today, and I wanna thank you for your testimony, are Ruth Sosa Martinez from the senior policy strategist with, Power California Action, and Monica Madrid, policy advocate with the Coalition for Humane Immigrant Rights, or CHIRA. Both witnesses are also economic justice fellows from the Solis Policy Institute with the Women's Foundation of California. And will tell a very compelling story that will demonstrate why this bill is all too necessary.

Welcome to you both and you'll each have two minutes. Okay.

Good afternoon, chair members. My name is Monica Madrid and I'm a state policy advocate with the Coalition for Humane Immigrant Rights, Chirla, and an immigrant economic justice fellow with the Solis Policy Institute within the Women's Foundation of California. I'm here today as a proud cosponsor of AB 2624 by Assembly member Bonta.

AB 2624 expands California's safe at home program to include immigrant serving organizations from and workers who face growing threats based on the communities that we serve. For organizations like CHIRLA, these threats and these threats are not theoretical.

We have experienced incident incidents where individuals were followed by staff followed our staff into the Sacramento and San Bernardino offices, because of their work survive supporting immigration communities. At our Los Angeles office, our headquarters, individuals have followed staff into the building and attempted to access, restricted areas. Our intake lines have received threatening messages and shown, and individuals have actually shown up to the house, house of family members of our executive director looking for our executive director.

They these instances are occurring during a time of heightened hostility toward immigrants and organizations that serve them. As immigrant serving organizations continue providing legal services, know your rights education, rapid response, staff are increasingly concerned that their personal information can be used to harass, intimidate, and harm them or their families.

AB 2624 provides a practical and proven solution by allowing eligible individuals to participate in the safe at home program. The bill helps protect home address information while preserving public accountability. No one should have to choose between serving their community and protecting their family's safety. California has long recognized that certain individuals face elevated risks because of their work, and anti immigrant and immigrant serving organizations deserve access to these same protections. We respect thank you, and we respectfully ask for your aye vote.

Thank you. Good afternoon, chair and members of the committee. My name is Ruth Sosa Martinez, Solis Policy Institute fellow, co sponsors of the bill. But I'm also the senior policy strategist at Power California Action. Power California Action serves young people and working families, and we have very deep roots in immigrant communities across the state.

As the daughter of immigrants and someone who works really closely with other organizations, many of us have become increasingly concerned about the safety of our volunteers, our staff, and our community leaders. For us, these concerns aren't hypothetical. In the past year, our own Board Members have been targeted, doxed, and stocked, for the work that they do in their communities.

We've had to become more intentional about how we, protect the privacy and security of our staff, of the youth leaders that we train and work with, and even of the, staff of other partner organizations that we work with, which diverts time, energy, and resources away from the core work that we, that we are doing. And we're not the only organizations who are taking these precautionary measures.

We're also concerned about the message that this sends to the next generation of generation of immigration advocates, attorneys, and organizers, service providers. We work closely with a lot of young people who either want to pursue careers or are already pursuing careers in public service and community advocacy.

But when they see that people in that field of work are getting harassed, targeted, and doxxed online simply for helping their communities, access critical resources and exercising their rights, these careers start to feel a lot more unsafe and unattainable for for our folks. Privacy isn't an abstract issue, and for many people, it's what allows them to continue showing up for their communities, without having to worry that their families might become targets. It's also what allows immigrant communities to feel safe seeking those resources.

And for these reasons, we respectfully ask your aye vote on 2624. Thanks.

Thank you both very much. Thanks for being here, and thanks for your testimony. Are there any other members of the public wishing to register your support for this bill? If so, please come to the stand up mic and share with us your name, your affiliation if you have one, and your position on the bill.

Good afternoon. Good-good afternoon, mister chair and members. Usama Muqaddam with the California chapter of the Council on American Islamic Relations in support. Thank you.

Good afternoon, chair and members. Karen Stout here on behalf of UNI DOS and Power California Action, both in support.

Thank you. Good afternoon. Jessica Hay with AFSCME California in support. Hi. Chloe Irmosillo with the California Immigrant Policy Center here in support. Also providing support on behalf of the Southeast Asia Resource Action Center and the Inland Coalition for Immigrant Justice. Thank you.

Good afternoon, chair and members. Cleo Bluthenthal with the California Community Foundation in support.

Alright. Thank you. Are there any witness- any witnesses in opposition? Or anyone that wishes to register opposition to the bill? Sign on, let's return it then to the committee for questions or comments.

Senator Gomez Reyes.

Thank you for bringing this bill. I think, Monica or one of you said that no one should have to choose between serving their community and staying safe. And I think this is one step that we can take to make sure that those who do serve the community this isn't supposed to be a a life and death job. But because of the sentiment against immigrants now, it has become that. And the harassment is not something that any of you should have to live with.

I think this is an appropriate bill to provide some protection. Thank you for that.

Senator Gonzales.

I do just wanna say thank you to the assembly member and the witnesses as well. I couldn't agree more with Senator Reyes about what is going on now in the state of our, affairs with, you know, the immigrant population and those that are advocating so many, day in, day out.

I've just talked to Angelica from from Chirla as well, and I I literally have to tell her and remind her, like, that we're here for her and as well as many of you because of just so much that you're putting, you know, you're putting your name, you're putting yourself out there, getting doxed. And it's just not it's not right. We have to take a stand.

This is one way to be able to cover you and I'm appreciative of that. That's why I signed up as a co author. So thank you again.

Senator Ochoa Bogh.

Thank you, mister chair. We appreciate the effort, that, you're bringing forth. I have actually supported, previous extensions of the safe at home program, in different, sectors. But I do have an issue with this particular one with regards to the private right of action that you have on here. And I think this, if I'm not mistaken and correct me if I'm wrong, but I don't think that the previous ones are, that we've extended have this particular component in there.

And that's one of the reasons why I can't support the bill today. And but I do appreciate the efforts. And I do believe just for the record that no one should be harassed at all. There shouldn't be that component. I've actually carried bills that had to do with right now, we had a bill that we have not been able to pass and it gets killed every every single year that we've introduced it with regards to removing personal addresses for people that do recall elections.

The signature and their address goes public. That it goes public and it, you know, we can't get that information off on there from from public. We can't get a bill to go through here so I am throwing that out there. But with regards to this, I completely agree with you. There shouldn't be anyone harassing anyone.

But the only concern I have is this particular private right of action on that. And if you'd like to make any comments on that, I'd appreciate that.

Yep. Thank you for the concern, Senator. This bill does not, contain a private right of action or if it does, I I would love to, help clarify that. It's actually the same language that was pulled from AB 82. So, which does not have that private right of action.

Yep. So that's a wrong file item number

It's 2624. AB 2624.

I will say that, there are penalties that are applied and it mirrors the language that is in AB 82. Penalties for exposing misconduct by itself. There's no language about so it imposes penalties when that are tied to intent to harm or threaten, applying only when there is a posting that is done with proven intent to cause harm or threats or someone ignores a valid safety based removal request that is backed by with evidence.

And I will also share to our senators here that this legislation will go through judiciary committee as well as public safety.

Alright. Are there any other questions or comments about this? I think the I mean, the private right is contained in the existing law that this program expands and kind of copies essentially. So it is there but it is and this says the as the author has noted. That particular issue is principally the jurisdiction or it is the jurisdiction of the committee on judiciary which will be the the committee of next hearing if it passes today.

Alright. It's been moved by Senator Padilla. Would you like to close?

Thank you. I appreciate the interaction that we've had on this piece of legislation. And again, I wanna thank our folks, Ruth Sosa Martinez, and Monica Madrid, who have come to testify yet again. I think this is your fourth hearing in combination on this piece of legislation.

At the end of the day, it's a very straightforward request to ensure that we provide safety for our immigration support providers now, when they need us the most to be able to do that and extends a very non controversial program that has existed for nearly thirty years to those who provide services to our most vulnerable communities in this moment right now and I respectfully request your aye vote.

Alright. We have a motion by Senator Padilla on AB 210 sorry. I'm already I've already turned the page. AB 2624 by Assemblywoman Bonta. Committee Secretary, would you please call the roll?

The motion is do passed to judiciary.

[roll call]

It's a 4-1 on call.

And the vote is 4-0 and we'll place that measure on call. Thank you very much.

Thank you.

Next we will proceed to AB 2103 by Assembly member Irwin. Welcome and please proceed when you're ready.

Well, good evening, chair and members. I'm pleased to present AB 2103. AB 2103 codifies Engage California as a permanent statewide public engagement program.

For too long, California has relied primarily on traditional forms of civic participation such as public comment, hearings, and written submissions. And while those tools are important, let's face it.

The majority of Californians don't show up to City Council Meetings or Senate privacy hearings. Engage California builds on a model that expands access to public participation by meeting people where they are. This program creates a structured process for Californians to receive balanced information, deliberate with one another, and provide input that can be translated into actionable recommendations for policy makers.

Here to testify alongside me is Jarrett Krumrei, the Lead Product Manager at the California Office of Data and Innovation.

Welcome and you'll have two minutes.

Thank you, chair. Good evening, chair. Good evening, committee members. My name is Jarrett Krumrei. I am the Lead Product Manager for the California Office of Data and Innovation and the Program Director for the Engaged California program.

The Office of Data and Innovation was established in 2023 as a standalone department to improve service delivery for the people of California through data, research, human centered design. It is the home of the Engaged California program.

The goal of the Engaged California program is to give residents a way to have meaningful dialogue with government. To shape policies and programs for topics that are important to them.

Our efforts to date have focused on topics of urgency and importance for Californians, including the LA fires recovery and rebuild efforts, state worker efficiencies to improve state services, and now around what government should do about the economic impacts of AI.

This is all part of California's work to introduce the practice of deliberative democracy as another tool for government to engage with its constituents. This is a movement that has been decades in the making and demonstrated effectively in places like Colorado, Paris, and Taiwan.

The Engage California program brings this work to California at a scale fit for our great state. I'm happy to answer any technical questions you have here today. Thank you.

Alright. Let's turn next. Are there anyone wishing to register their support for AB 2103? Please provide your name, affiliation if any, but not your lobbying firm, then we're not here for advertisements and your position on the bill.

Okay. Not not to advertise, but yeah. Robert Booking with TechNet and up in support. Thank you.

Charles Contrebecki, intern on advocacy on behalf of Elevate California in support.

Alright. Is there are there any witnesses in opposition?

And or does anyone wish to register their opposition?

Seeing none, we'll turn to the committee for any questions or comments. Senator Mcnerney?

I thank you, and I thank the author for bringing this forward. I understand this authorizes the legislature to fund the program but doesn't provide funding.

The funding needs to be approved by the legislature.

Are there would there be other sources of funding that would that it could accept?

I don't know if you have the answer to that.

I'm here to talk technical aspects of the program today but can certainly work with the assemblymember to get back to you.

So I mean if a a philanthropist wants to fund this then that would probably be acceptable. That's my question.

Well, we will make sure that is addressed in the bill. That certainly makes sense because it is a program very important right now that Californians are able to voice their opinions on these big issues.

So many people feel disconnected from government. And again, city council meetings and letters to the editor are used by very few folks. So this is a way to engage many more Californians in the and let them inform the policies that we're looking at in this legislature.

So many agree with that and it's good to see TechNet in support.

Yeah. That was I was so not used to being in support that he came up in opposition.

Alright. And I will yield back to this year.

Alright. Thank you. Senator Ochoa Bogh.

Hi. Thank you, Senator Cabaldon. So questions, I appreciate the intent of the bill with trying to get more information into public, into the public sphere. Let's make sure I have I'm in the right bill. I wasn't the right bill by the way, previous one.

Really, really appreciate that. I think where I have concern is right now, we have here that it would authorize the Speaker of the Assembly, the President pro tem, Secretary of Governmental Operations to select the deliberation topics.

And then requires the department of or within the Office of Data and Innovation to recruit participants, administer deliberative engagement platforms, coordinate with state agencies, report on program activities and so forth.

I think the concern I have is trying to bring a little bit of balance so it's not constructed in a way that becomes very party. I'm looking into more of a balanced approach as far as the topics and be able to address those issues that you have right here.

Because if right now, in the case as a super minority republican, in a state legislature that is a super majority, you know, Democrat. When we're looking at at these folks who are going to be deciding and having decision making as to recruiting the the participants, the deliberation of topics and so forth. It's gonna be heavy on one perspective and it won't have a balanced perspective.

Have you considered have there been any consideration of how to make it a little bit more bipartisan in order to have a more objective approach in addressing some of these issues and deliberation of issues and including the participants out here?

We can certainly look at that. But I think if you look at the issues that Engage California has explored so far, it's issues like the California fires and AI, and those are very nonpartisan issues. So we would certainly hope that continues. And if we can, you know, certainly look at a way to make sure that is codified.

It's just, you know, so I was looking at my in my ideal world, you know, you'd have the vice chair of whatever, super minority party would be in place in both houses would be a great balance, I think, in trying to deliberate and have conversations on that.

Just in my perspective on there, I always try to look to see how we can do things a little bit more, especially when you're looking at codifying an Engage California program that established a framework for a structured public deliberated intended to facilitate dialogue between California's and state government, regarding and I'll give you an example. I sent out an email on a weekly basis where we're having committees and I don't give my personal opinion.

I just let them know here's the bills are coming up in committee. These are the authors and this is a couple of questions and here's a link to the actual bill language so that you can have access to it.

So I try to be as nonpartisan as I can in delivering information to my constituents. And I think it would be great to see that if we're going to be engaging in such a place, we make it nonpartisan as neutral as possible or balanced.

Well, like I said, I think the topics that have been picked by the Governor's office so far have have been topics that are important to the community, whether they are Republican or Democrat. But maybe you can let us know what is the process that you are currently going through.

Absolutely. So first of all, I appreciate the acknowledgment in terms of transparency being a core principle of our program. And by design, any engagement that we embark upon, our intent is to reach all members of the audience regardless of demographic, political affiliation, etcetera.

To date, as Assemblymember Irwin had mentioned, the focus has largely been on issues that are of timely importance to Californians, in concert with, our partners across the state, with the Governor's office directly.

For instance, with the state employees engagement, to work with state employees to understand ways that we, as California residents and as workers can improve our efficiencies and operations within the state to deliver better services.

Working with all department heads and secretaries and under secretaries and their staff to reach every department team through their channels and mediums. And the last thing I would say in terms of of the output of all of our engagements, going back to our transparency principle that we publish all results of our findings.

That includes the raw data, anonymized data from participants so that any third parties can independently analyze our results and our methodology on their own. That includes academia, civil society and different political affiliations.

I absolutely appreciate that as well. And just as a further, clarification just for the record and for the public. When we're talking about departments, department heads and agencies, many of them are appointed by our Governors. So it also gives one perspective usually on that end. So I appreciate the effort to try to be as transparent in it, but in a state where we have literally everything dominated for the past sixteen years, whether it's departments, agencies, the legislature, the governor.

I'm trying to find, you know, where we could find places. And it would be the same way for me if it were reversed, if the Republicans had a super majority. So freeing that balanced perspective so that we're not working in constantly in silos and echo chambers.

And that's I think the biggest and that's why my priority is to ensure that we have a different voice that comes in on that to bring a little more balance and where we avoid those silos and echo chambers.

Senator Gomez Reyes.

Thank you, mister chair. When I see something that is going to provide civic engagement, public participation, this is what we want. Working in silos, that's us. We're here. We're here at Sacramento and often times unless we have those town hall meetings or unless we send out the letters to our constituents, we're not going to hear from them.

I think this is a great idea to be able to get people involved. And I would hope that everybody would be involved. And if, just as here, if the majority feel one way, the minority will always speak. And sometimes the minority opinion is something that may sway the majority, in making a decision or to add an amendment to make sure that we do something, that includes that voice. But I think public participation and providing information.

Just and I do wanna be very specific have a specific question. Is this nonpartisan?

Well, that is absolutely the intent. And again, this is a codifying a program that's been going on for the last year or so. And all the, you know, what we have seen so far is that we have gotten a lot of good information from a very broad sector of our constituents. And that's something, again, I'll repeat that we we don't necessarily see here in the hearing room.

Thank you. With that, I would move the bill.

Alright. Thank you. So moved by Senator Reyes. Are there other questions or comments?

I'm ecstatic to see this bill.

Facebook is attracts my every move over many many years reminded me this morning that on this day, ten years ago, I was traveling giving speeches in Korea, England, Spain, and Qatar on this very topic of what is the present and the future of civic engagement. What you mentioned at the outset of the at the hearing that, you know, not all not most of California is not here attending our privacy committee hearings.

But of course, the if you really think about it, well, what if they did? Like what if we got what we wanted and everybody showed up at all the hearings that we announced? Like how would we even do business?

Like how would we get actually take their feedback and engage them and do anything with them? These are systems that are sort of the systems we've set up since the 50s and the 60s are designed to not draw people to participate.

We can only we want we say we want as many people as possible but we can really only handle a very small set. So we need to be like experimenting and designing new things and so I really appreciate this program.

It's been funded in the budget subcommittee.

It's also had rave reviews. You know, we've been supporting ODI for this purpose. It's not the topics are nonpartisan. The process is nonpartisan. A Senator from any party is eligible to be elected by her or his colleagues to be the president pro tem.

That's not a partisan role. And really, these are not the point here is not a process that just picks the most interesting topics, the most interesting issues. Because what engaged California is about is about figuring out how to evolve and improve our democracy itself.

And so not every topic, there may be topics that are really interesting but not suited for this. Topics that are really important but not suited for this.

And so that's why the process that's outlined here is designed to pick things, pick topic areas. Yes, we will learn and make better decisions from but also they will give us insight about what will excite people to participate, what will make them feel and for real be more effective at that participation as well. So I think this is a fantastic program. It does need to be in statute and it has been achieving budget support in the last couple of rounds.

So far the legislature and governor have been committed to it and thank the author for carrying this.

Alright. So we have a motion from Senator Gomez Reyes and committee assistant. Please call the roll.

Did you close?

Pardon?

Did you close?

Oh, sorry.

Would you like to close?

Like my

So sorry.

I'm just respecfully asking for aye vote.

Then let's call the roll.

Okay. The motion is do passed to appropriations.

[roll call].

Six to zero.

And the vote is six to zero. We'll place that bill on call. Thank you.

Thank

Thank you. And next, I see Assemblymember Lowenthal is here. So we're gonna begin, here's a couple bills. We'll begin with AB 2. Welcome, Assemblymember Lowenthal and begin whenever you're ready.

you very much.

Thank you, mister chair. Thank you, members. Thank you, senators. A sneaking suspicion this won't be the only time I come before this committee. I'm so pleased to present AB 2, which will hold social media platforms accountable for the harm they cause children and teenagers.

This legislation would impose financial responsibility by setting appropriate damage levels on large social media companies, if and only if their negligence has been proven in court. Those of you who are parents, when your child puts on a bicycle helmet, when they get in a car seat, when they play with toys, aren't you confident that the safety standards not only come first, but have been rigorously tested and transparently reported on by manufacturers?

I am because I know the duty of care standards are bedrock in the product development cycle of everything our kids use. Everything that is, except for the one product they use the most. The product that they're spending on average five hours a day on.

The one product that is empirically having disastrous impacts on their mental health and development. That product is social media. This is why AB 2 is so critically important. It holds social media platforms to the same legal accountability standards as any other company for harm they cause to children and teenagers, and assigns appropriate damage levels for the harm that's caused.

The bill sets damage levels that can only be meted out if a court imposes financial responsibility on large social media companies when their own negligence has been proven.

This bill is not new. It's my third year working on this policy. And I find myself compelled as a parent and as a legislator to continue to work on this piece of legislation because we are truly struggling to hold the platforms accountable through regulation. And the platforms are still not taking this problem seriously. Children's parents, researchers, doctors, regulators, and lawmakers alike have clearly delineated the problem.

Social media platforms are causing unhome, untold harm to our kids. And what we have known to be true has been reinforced by numerous studies and exposed internal documents from the platforms themselves. These children have been exposed to a product that by its design is addictive and has resulted in harmful and disastrous outcomes for child users. But our pleas for social media platforms to self regulate have gone unanswered. Our attempts to regulate the platforms are being challenged by the platforms and their trade associations in the courts.

And the public is demanding action, and their outcry is growing stronger by the day. Unfortunately, social media platforms continue to deny that their algorithms and design features are contributing to these harms. Even in the face of two Bellwether rulings earlier this year, which found that social media apps should be treated as defective products for being engineered to exploit the developmental- developing brains of kids and teenagers, and did not adequately protect users. The platforms has refused to change their designs, which is harming and killing kids.

So why is this approach to AB- why is the approach in AB 2 necessary?

Because the public is calling for social media platforms to issue a rea, a product recall. To make sure that their product that we have, like it or not, come to depend on and use every day is designed to be as safe as possible for us as consumers. That should go without saying when it comes to our children. But the platforms are continuing to put their profits before everyone's safety and well-being, and that just needs to change.

If the platform's profits are more important to them than anything else, it's crucial that we have the right financial incentives in place to ensure that they take this problem seriously, that they implement real change, and they simply do not chalk it up to the cost of doing business.

And that is precisely what AB 2 does. It signs appropriate damage levels to ensure that when a large social media platform's own negligence is proven in court for their failure to exercise ordinary care, the damage levels are sufficient to ensure that they are incentivized to change.

During arguments on the punitive damages in the KGM trial, Mark Lanier, the lead lawyer for the plaintiff, was reported to have shown the jury a jar full of M and M's during the argument on punitive damage, noting that each piece, each piece of candy represented a billion dollars of the company's worth. He said you can take out a handful and not make a difference. He said scooping out a few with his hand, you can take out two handfuls and not make a difference.

The jurors concluded that Meta and Google should pay 3,000,000 in compensatory damages and additional 3,000,000 in punitive damages in the KGM trial. That verdict alone has not incentivized the platforms to heed the call of the public to change. But it does represent a critically important opening move, and there are more cases to come. AB 2 will serve as an essential guide to the courts and juries when these cases come to trial, to assess appropriate damages.

To ensure the platforms are taking these cases and the public's call to put our safety first seriously.

You might be asking yourselves, why is AB 2 necessary if California joins the growing global movement to establish a minimum age for social media use? Well the short answer is that we have no guarantee that the platforms will comply with the law. And there's already evidence to suggest that they're not doing their part to enforce minimum age requirements in other countries. So even if we were to establish a minimum wage requirement, the platforms do not abide by the law.

We will still need to ensure that they are assessed appropriate damage levels for the harm that they are causing.

And in some, social media platforms are not exercising ordinary care. They are fully capable of designing algorithms and features that promote safety and not addiction, But without facing sufficient financial liabilities for their failure to exercise ordinary care, they are refusing to change. And they're continuing to drive children and teens to harmful content, encouraging them to engage in dangerous and sometimes life threatening behavior. AB 2 does not change California's underlying law or the burden of proof required in court.

AB 2 applies appropriate financial incentives and accountability to prompt just a handful of companies who are earning enormous profits off of children and teen users to be more careful and responsible partners, especially when it comes to our kids.

Let me be clear. Financial awards, no matter how great, will never be able to compensate for the loss of a child. Never. But that's how dire the situation is. It is literally life and death in far too many of these cases.

But incentivizing change, protecting other kids from the tragic outcomes experienced by far too many families in California and around the world, that is what the public and many of the impacted families are calling for. Make your product safe. Design it like your kids are using it. Don't let another kid become a casualty of social media. That is why we need AB 2.

I'm so pleased, mister chair, to be joined by Juliana Arlen Arnold with Parents Rise and Nicole Rocha with Tech Oversight California who are here to testify in support of the bill.

Welcome to you both. And you'll each have two minutes.

Thank you. Good afternoon, Chair Cabithon and committee members. My name is Juliana Arnold and I'm the executive director and cofounder of Parents Rise, a national advocacy organization led by survivor parents. And I strongly support AB 2. More importantly, I'm a mother of two daughters.

My eldest Olivia just graduated from college. Coco, my younger daughter who was clever and curious beyond her years, never got that chance and never will. At 17, she was groomed on Instagram, lured to meet an older man, and given a counterfeit pill containing fentanyl. She died in 2022. There are no words for bearing your child, only the silence that is left behind.

Before Coco died, we watched social media take more and more of her from us. It stole her sleep, strained her mental health, and made it harder for her to step away. We set limits. We used parental controls. We did everything we were told to do.

It was like playing whack a mole. And yet, the product kept pulling her back and the parental controls just didn't work. They did not suffice. And it leaves parents feeling helpless. Parents should never be feeling that way when they're trying to make sure that their kids are being are safe.

So Coco's story is uniquely ours, but the pattern is heartbreakingly familiar. Children have been targeted, exploited, and pulled deeper into danger by products designed to keep them engaged even when the harm was foreseeable. We now know it is not ignorance. Through the trials, court records and testimony have shown these companies understood the risks to children and failed to act. They knew children paid the price, but they chose profits over our children.

AB 2 applies a basic rule that governs every California business. Exercise ordinary care and and do not cause foreseeable harm. AB 2 is not about punishing platforms for what someone says. It's about the company's own choices. How they design their products, drive compulsive use, and fail to act when harm is clear.

I have to ask

you to What was that?

I'm gonna have to ask you to come to a conclusion.

Okay. Can I put it in the conclusion

just for one second? Yes.

Thank you. For these companies, it's another cost of doing business. Accountability will not bring my daughter back, but it can help spare another family this nightmare. Our children deserved ordinary care, and they still do. On behalf of Parents Rise and the children who no longer can speak for themselves, I respectively ask for your eye on vote on AB 2.

Please do not let more families join ours.

Thank you. Alright. Thank you. You're welcome.

Good afternoon. My name is Nicole Rocha and I'm here on behalf of Tech Oversight California. TechOversight is an advocacy organization that champions meaningful tech accountability reform. This is the third year that I personally have testified on this important measure. What sets AB 2 apart from legislation of prior years is that we have recently seen the first monetary verdicts in American history holding social media companies liable for design driven harm to users.

These cases have began to flesh out the duty of care that was questioned by committees in prior years. Tech oversight worked closely on the Los Angeles Bellwether case involving thousands of plaintiffs, families, school districts, states, and localities in which Meta and Google were found liable for designing platforms child. The day before, a New Mexico jury ordered Meta to pay 375,000,000 in civil penalties for misleading the public about

platform safety and enabling child sexual exploitation. These verdicts

con confirm what parents and exploitation. These verdicts con confirm what parents and advocates have known for years. The harms to youth are rooted in product design and companies have known the harm that their products are are inflicting. So why is AB 2 needed in light of these verdicts? Simply put, statutory damages will allow for more consistent results where all children's lives and experiences are similarly valued.

In many cases, it will save families from the pain of establishing the worth of their child's life. Statutory damages will also create a clear path forward for companies to evaluate exposure and liability, with the state of California clearly signaling that all children deserve a safe and healthy online environment. AB 2 is a common sense measure that is long overdue. In the name of consistency and fairness, we urge your aye vote.

Thank you both. Does anyone wish to register their support for AB 2? Please come forward and provide your name, your affiliation, and your relation. Sure.

Good afternoon. Good evening, mister chair members. Christopher Sanchez here on behalf of the Consumer Federation of California in support.

Greetings, mister chair and members. Pamela Gibbs representing the Los Angeles County Office of Education, and we're proud to join the author and witnesses in support of the bill.

Mister chair members, Amy Brown on behalf of the California Charter Schools Association in support.

Cheryl Westmont with Mothers Against Media Addiction in strong support. California.

Diana Hawkins in representing MAMA Silicon Valley in support.

Good afternoon, mister chairman, members. Ed Howard on behalf of the Children's Advocacy Institute at the University of San Diego School of Law in strong support.

Mike Eofy on behalf of Common Sense Media. Proud to sponsor this bill and support.

Alright. Thank you. Are there any lead witnesses in opposition to AP two?

Alright. Thank you. Are there any lead witnesses in opposition to AP two?

Welcome to you both and you at least have two minutes as well.

Alright. Good evening, Mister Chair and Senators. Dylan Hoffman today on behalf of TechNet and we are opposed to AB2. And while we completely agree with the intent, of the bill and our platforms work diligently to, constantly improve and protect their users, we believe the bill will do very little to actually solve, any of the problems identified. One of the major issue with this bill is the misconception around distinguishing between content and conduct.

We believe the line isn't so easily drawn between content and design, particularly when it comes to content serving features like algorithms, recommendations, and even direct messaging. This bill has been talked about as purely about product liability. However, many cases and examples that are cited of social media causing harm are rooted in harmful content, not the platform design. And even then, these cases are incredibly fact specific and individualized.

For example, a social media site dedicated to, cute animal videos or puppies, wouldn't cause an eating disorder, even if if it used the exact same recommendation algorithm, and features as a major social media platform.

It's the content, not the feature, that is the issue. And courts are currently considering this distinction and in which instances plaintiffs should be able to recover damages. Not only are there hundreds of cases pending that would be impacted midstream, but the ninth circuit is also considering arguments about whether content serving features and platforms editorial discretion have first amendment protections. The fact specific and highly individualized nature of this issue dramatically undercuts any incentive, this bill is trying to create.

And it's, the bottom line is plaintiffs can already sue social media platforms.

Courts have shown they're willing to let them proceed in spite of Section two thirty, and juries are willing to award significant damages and penalties. Damages are rarely an issue in in these cases, and plaintiffs are already able to be be made whole. And if anything, these cases show that this bill is unnecessary. Additionally, the bill is somewhat inconsistent on whether the intent is to punish egregious intentional conduct.

The findings and declarations even reference the need to ensure that platforms that are knowingly causing injuries should receive heightened penalties.

But the bill isn't actually limited to those most severe instances. A prior version of this bill, AB 3172, was amended to require conduct to be knowing and willful before the bill's penalties were triggered. Those amendments also limited these heightened penalties to public prosecutors who are more likely to bring cases that are more narrowly limited to product design and features and not to content.

Bill, I'm gonna ask you to bring that come to a conclusion.

Last line. Because this bill will substantially and then unnecessarily increase litigation and penalties without actually solving a larger problem, we must respectfully oppose AB 2.

Thank you. Alright. Thank you.

Chair, Members of the Committee, in interest of time, I'm, I'm Aidan Downey, with the Computer Communications Industry Association. I wanna echo what Dylan said. I don't wanna rehash anything. So stand ready for questions.

Thank you.

Perfect. Alright. Does anyone wish to register their opposition to AB2? If so, please come forward to the microphone and provide your name and affiliation and position on the bill.

Ronak Dalami on behalf of Cal Chamber also in opposition. Thank you.

Catherine Charles on behalf of the Chamber of Progress in respectful opposition.

Annalee Augustine with the Civil Justice Association of California also respectfully opposed. Thank you.

Danny Canto Kaizer on behalf of the Electronic Frontier Foundation respectfully still opposed.

Alright. There's no further testimony. First, thank you to everyone who participated. And let's turn to the committee for questions or comments. Senator Padilla.

Thank you very much, Mister Chairman. I thank the author more comment. First, I wanna thank you for bringing the bill. I know it hasn't been easy. Now the author and many of my colleagues are all still aware that I do a substantial amount of work in this space in my own portfolio.

Understand the intent here and have read and reread the analysis. He went back and took a look at the language a couple more times to make sure I wasn't either seeing things or hearing things. I think fundamentally for me, Mr. Chairman and colleagues, bill boils down to one thing and that is the consumer protection angle here as a matter of policy consideration for this committee. And fundamentally consumer protection is only real and is only appropriate when there is certain and meaningful enforcement.

And enhancing remedies without creating a new cause of action, without changing standards. Question for this member is, is this bill enhancing consumer protection by virtue of enhancing remedies and is that an appropriate policy approach for this legislature to take along those lines? I think the answer to that is pretty overwhelmingly clear when you look at the evidence.

The questions of the sufficiency or deficiency of those remedies, what any impact they may have, potential arguments someone make in defense, those are questions for the committee on the judiciary and perhaps a a trial court or a jury to look at in a particular case or cause of action. But I think this bill is appropriate.

I think it's pretty clear what the bill is and what the bill isn't. And so I wanna thank the author again and mister chairman at the appropriate time, I'm happy to move the bill.

Senator McNerney? I thank the Chair. I thank the author for bringing this forward. My mind is common sense. If there's things that if there's something out there that's harming children, it ought to be taken care of, but it'll be dealt with.

And this is one way to do that. Protecting children is is maybe the most important moral imperative of any society. Now, algorithms, can be addictive. They can be addictive like, addictive drugs. And they can they're designed specifically to be addictive.

They're designed to get that extra click because every click means more money. So I just people are growing more and more distrustful of big tech. They're growing more and more distrustful of artificial intelligence. And it's just common sense to to put guardrails out there that'll provide basic protections. And then, of course, these companies, the, M and M metaphor is certainly appropriate.

They can afford, and if we wanna make them stop, then you gotta put enough pain to make them stop. So anyway, I'm gonna support this bill.

Alright. Senator Gonzales.

Yeah. I too wanna

thank you, Assemblymember. You and I have spoke about this a lot having three girls, three boys at home, and, we know how how difficult it is as a parent. And I, I want to thank your witness for sharing her story. I'm so sorry for your loss. And I agree with my colleagues here.

I think it's pretty pointed that you're addressing the remedies. There is no you know, it's it's one thing to put forward the the the the guardrails and ensure that there are regulations, on social media, but the other part is the enforcement.

And there's also recent precedent as well that showed us that there is a road map to ensuring that there are there are enforcement mechanisms that will actually, hopefully, get to, less children being affected online, less children, being lured, less children being overall impacted in the state of California. With that, I support. Thank you.

Any further questions or comments? Parts moved by Senator. But the I'm gonna support the bill today. I, I, I share the the the feelings of a lot of the members of the committee on this issue. I have I do have some questions about about it, but they're they're entirely within jurisdiction of the Judiciary Committee which I'm sure will give it its full attention.

So I'm not making a permanent commitment on all the way through but I think the cases have been made. And, with that we do have a motion. So, Committee Assistant, would you please call the roll?

I'm gonna close.

Oh, sorry. I do I do this every time. Assembly or Lowenthal, would you like to close?

Mister Chair, I just want to thank you and the Members. I want to thank you and your staff and the committee staff for the time and the meetings that we had on this topic. You know, I just want to opine to everyone right now. What does it mean to to actually stand up for consumer protection in the digital age? What does it mean when you have multi trillion dollar companies that scale the entire span of the globe.

What does it mean when you're trying to regulate entities that are larger than the state of California itself? There has to be a bite to that for them to be able to listen. I think that was articulated by the members of this committee, but it is something that we all need to be asking ourselves. What is the right amount? How do we get them to listen?

And beyond that, one of the most amazing things about AB2 is that it doesn't change the law at all. And so it can go into effect immediately. And while the platforms continue to fight and rage against the vast, vast majority of Californians who are saying we want regulation. We want to start from a place of safety and work backwards from there. This bill can go into effect immediately.

And while they take that stance in court, and why it can take years and years for those proposed laws to go into effect. We can provide remedy for millions of children in the state of California and be a model for elsewhere, and hopefully require them to design their products from a place of safety. And with that, I respectfully ask for your aye vote.

Alright. Please call the roll.

[Roll call]

Both five to zero. I'll place that measure on call. Next we're gonna proceed to AB 883. Assemblymember, Lowenthal. And thank you so much for being with us.

And so Assemblymember Lowenthal, AB 883.

Thank you senators for the opportunity to present AB 883. I'd like to begin by thanking the committee staff for their thoughtful work on this measure. I will be accepting the committee amendments. Now as elected officials, it should come to no surprise to any one of us that, threats against public servants are increasing across the the country. Political violence is not hypothetical.

It is a growing reality. According to a 2024 report from the University of San Diego, 66% of all elected officials reported being on the receiving end of threats and harassment. I'm in that group. I know there are people on the dais that are also in that group. Even more, forty six percent of women, thirty nine percent of men have considered leaving public service as a direct result of the threats and harassment that they have experienced.

National data paints a similarly troubling picture. More than forty percent of state legislators have experienced threats or attacks within the past three years, and nearly ninety percent reporting experiencing harassment, stalking, or intimidation. Judges are increasingly facing similar dangers. The US Marshals Service reports that threats against federal judges more than doubled between 2021 and 2025, while thousands of threats against state judges have been documented nationwide.

In 2020, Daniel Anduril, the son of a federal judge in New Jersey, was murdered at his family's home by an individual who had targeted the judge.

More recently, elected officials and their families have been victims of targeted violence, underscoring the very real dangers facing those who serve the public. As evolving technology has expanded the ability to collect and distribute personal information online, data brokers have made it easier than ever for bad actors to locate individuals and their families. The Federal Trade Commission has found that data brokers collect and maintain extensive personal information on nearly every American household, often retaining that information indefinitely.

But California has already taken a national leadership role through the Delete Act and the creation of the Delete, Request, and Opt Out platform known as DROP. Since the platform became available earlier this year, more than 300,000 Californians have already signed up to exercise greater control over their personal information.

The amendments before you today strengthen this approach by addressing concerns raised by stakeholders, while maintaining the bill's core objective of protecting privacy and safety. Under the amendment bill, Cal Privacy will create informational materials detailing how elected officials and judges can utilize the DROP system to remove their information from data brokers. This information will be distributed through the Secretary of State, through local filing officers, and the Judicial Council to ensure those serving the public are aware of and able to access these important protections.

Additionally, rather than creating a separate process for a limited group of individuals, AB 883 will reduce the deletion timeline through drop from forty five days to thirty days for all Californians. This approach improves privacy protections equitably, ensuring every Californian benefits from faster deletion requests, while also increasing awareness among elected officials and judges about the tools already available to help protect themselves and their families.

Public service should not come with the expectation that individuals or their loved ones will face threats facilitated by the widespread availability of their personal information. AB 883 takes a practical balanced approach that strengthens Californian's existing privacy framework, improves access to important consumer tools, and helps protect those who have answered the call to serve their communities. Very pleased to be joined today by Doug Subers, representing Californians for Consumer Privacy, who is here to testify in support of the bill.

Thank you, mister Chair and Senators. Doug Subers on behalf of Californians for Consumer Privacy. Pleased to sponsor AB 83 and strong support the measure. We'd like to thank the author and his team for their leadership on this issue and all and all the work thus far, and also thank, you, mister chair, and the committee, for all your work in arriving in amendments.

All told, Californians for Consumer Privacy is really focused on strengthening privacy rights and improving mechanisms for Californians to exercise those rights, and we think this bill, meets that test in both regards.

As noted in the analysis and discussed by the author, we've seen a startling increase in threats and violence towards elected officials including the horrific incidents in New Jersey and Minnesota, and and incidents here in California. AB 883 will ensure elected officials and judges get information about the ability to exercise their rights under Cal Privacy's DROP system.

And our hope is that their better education and use of these tools will help limit the proliferation of personal information that we can use, could be used by someone to exercise some type of harm towards an elected official. Ultimately, better safety for elected officials and their family will promote or continue to promote the best and brightest of our society pursuing roles in elected office or or in the judicial service.

As noted by the author, pleased to work with a variety of stakeholders, to, address concerns and believe this, bill meaningfully improves rights for all Californians.

For those reasons, we respectfully ask for your aye vote. I'd also just like to register support for DeleteMe, a service provider company that helps Californians exercise their rights. Thank you.

Alright. Thank you. Does anyone wish to register support on AB 883? If so, please come to the microphone and share with us your name, affiliation, if any, and your position.

Maureen Mahoney on behalf of Cal Privacy. The agency has a supportive amended position on the bill in print and we very much appreciate the committee amendments as described in the analysis. Thank the committee and author.

Okay. Are there any lead witnesses in opposition?

Just delivering brief comments. Good evening Chair, Members. My name is Arba Wojciech with TechNet. We actually removed our opposition to the bill in print due to the author taking several amend several amendments regarding clarity of duties and explicit reference to permitted use purposes. That said, even though the timeline was short and the PRA still remained, our position was removed due to understanding that the population of AB 883 was narrowly scoped to individuals addressed in the bill, elected officials and judges.

Regarding the proposed amendment analysis with a shortened timeline to delete from 45 to 30 plus the expansion of what PI can be deleted pursuant to SB 923 passed by this committee earlier this year. If the PRA remains in the bill and is subsequently applied to the entire delete act, not just limited to elected officials and judges, we may need to return to an imposed position.

But we do look forward to seeing what the actual language would look like being proposed and working with the author to address any remaining concerns. Thank you.

Thank you.

Good evening. Ronak Dalami on behalf of Cal Chamber. Our letter somehow wasn't, reflected in the analysis. We are still in an opposed unless amended position. We will take a look at the amendments when they are in print.

But with the, PRA, I think we will still be opposed. Thank you.

Clayton Downer with the Computer and Communications Industry Association. I'd like to echo Carl Chambers comments in respectful oppose unless amended position.

Good evening, Chair, Members. My name is Tiffany Barbee on behalf of the ACLU Cal Action. We wanna thank the committee and the author for the amendments addressed in the analysis. And should those amendments go into print, we would be happy to drop our oppose unless amended. Thank you.

Alright. Seeing no other witnesses, let's turn back to the committee for questions or comments. Senator McNerney.

I thank the Chair and I thank the author for bringing this forward. As he mentioned, many of us on the diocese have been subject to threats and other things. And it's unfortunate that we're in this situation, but there's so much vitriol on on this social media So we came in earlier discussion that this is necessary. I'm gonna have to support this. And I'll yield back to the Chair.

Alright. Further questions or comments?

I just want to thank the the author and the sponsor on on the on the bill for for the work on on the amendments. And we're trying to do a couple things here, obviously. And I mean, first and foremost is to is to promote safety for folks who are stepping up to serve their communities and their and democracy in various ways.

And we've seen the the rise and the intensity, the severity of political violence throughout the country, but certainly here in California felt especially by by, women who are serving in public office, LGBT officers, but everyone, as well. And also understanding that we've had a lot of the legislation we've gotten, to improve the Delete Act and other piece of legislation is because we as elected officials live in the same world as everybody else.

And so when when when we have to go fill out the the drop back the drop request, your CalPERS are like, hey, this this is kinda hard. It's not. Just FYI, it's very easy. But when we do experience those, we make we make better policy. And and so we we we do try where we can to not wall ourselves off and create, you know, special pass throughs.

Both for fairness, but mostly so that we can experience the California that we are legislating about. This bill with the meta version absolutely walks that line appropriately and not without sacrificing or compromising on safety in any way. And also appreciate that for all Californians, the provision that was in a bill of mine is now here to shorten the the the the the lead act timeline now that we have enough experience with the act. We know it's technically feasible to implement in that way.

So with that, we have a motion by Senator McNerney.

Would you like to close?

Respectfully ask for your vote.

Thank you. Alright. Please call the roll.

The motion is do pass as amended to judiciary. Roll Call.

Alright. The votes four to zero. We'll place that measure on call. We are our remaining bills are all by a single author.

So we're going to await her arrival and in the meantime, the committee will stand in recess.

First, the, privacy, digital technologies, and consumer protection committee will reconvene in thirty seconds. Alright.

We'll reconvene the Senate Committee on Privacy Digital Technologies and Consumer Protection. Wanted to note that, at the request of the author, AB 2564 will be put over to our next hearing.

And now we'll proceed to AB 2023 by Assemblymember Wicks and Friends.

Yes. Friends.

Welcome.

Assembly Member of Bauer Kahan, yes. Thank you, Mr. Chair, and I want to thank the committee staff for all of their hard work. And I, we accept the amendments, the committee amendments.

So this is a chatbot regulation bill. And I know, Mister Padilla here who serves on this committee has been such a leader in this space. And we are very happy to be working collaboratively with you, Sir, as well as the chair, and many of our colleagues, and Miss Bauer Kahan obviously had a bill in this space last year as well. So a lot of us have been going at this trying to figure out how do we keep our kids safer online.

We've gone through the some of which have been horror stories with regard to social media and trying to make sure our kids are safe when when it comes to social media.

And now we are looking at chat bots, which can be a very effective tools for our children, but also we need some common sense guardrails and safety standards put forth for these chat bots. And the reason for that, 72% of teens have used an AI chatbot. More than half are regular users. 33% of teens use AI companions for social interactions and relationships.

In addition, research shows that children are much more likely to view AI chatbots as quasi human and to trust them more than adults, actually, and to seek guidance from them, as opposed to adults in their life who are trying to steer them in the right direction.

There are a lot of risks to this new technology, especially prolonged interactions with chatbots can be particularly risky. As our kids are still developing and growing, these interactions with chatbots have risks and risks lead to physical harm and negatively impact their mental health. And we have seen how dangerous chatbots, have become in terms of lost life, actually with regard to our young people.

So AB 2023 builds on the great work of Mister Padilla, Miss Bauer Kahan and creates a comprehensive framework for regulating chatbots for children by one, requiring an age verification of users. This is built on the premise of the age signal bill that I passed last year that's going to affect January 1 of next year.

Risk assessments, safeguards including default settings, parental controls, noticing, and measures to prevent suicidal ideation and self harm, prohibitions against targeted advertisements to children, and selling or sharing their information, and then third party audits for compliance.

These are the basic guardrails we think we need. With here, we have Nicole Rocha who's gonna be, here to testify on behalf of Children Now. But before we get to that, I want to offer the floor to my dear friend and colleague, Miss Bauer Kahan.

Thank you. And I just wanna reiterate, my gratitude to everyone here. Senator Padilla, who's obviously worked so deeply on the chatbot issue. Senator McNerney, who's working on standards that will be a fundamental piece to the third party verification elements in this bill.

And obviously, Assembly member Wicks on her age signal without which we couldn't protect kids online, I believe at all. And this is so critically important because of the moms that sit here.

And I know you all heard from Maria Raine when Senator Padilla's version of this bill was up in this committee and her story of her son and what happened when he was drawn in by a sycophantic chatbot.

And that hits home as a mother of a teenage son who uses chatbots to help with advanced math problems that mom can no longer help with, I understand why our kids are using these tools and why they're really beneficial.

And that's exactly how Adam started when he began his use of a chatbot. And then through the sycophantic behavior of the chatbot, it drew him in until ultimately he died by suicide. And so we should be able to give our kids a chatbot to help them with math and they should have the educational benefits of it.

And we should also know that we are handing our kids a tool that has been tested to be safe. That will not tell them how to take their own life. Will not them how to harm themselves. And those principles are frankly fairly simple, and this bill goes a long way and will be one of the strongest in the nation, if and I guess I'll just say when it gets signed into law as I speak it into being.

Although I will note that New York did just pass unanimously a law on chat bots that continues this incredible work.

So we are, moving alongside our fellow states and ensuring that these are tested and safe for children by design. And with that, I will turn it over to Nicole.

Welcome. And you'll have two minutes.

Good evening, chair members. I am Nicole Rocha here on behalf of Children Now. Children Now takes a whole child approach to improving the lives of California kids and works across health, education, early childhood, and foster care to ensure all kids have the supports that they need to survive sorry, to thrive. After many years of fighting for legislation to keep kids safe online, youth and families finally have the wind in their sails.

Earlier this year, we saw verdicts totaling hundreds of millions of dollars with juries finding that social media intentionally addicted children and failed to protect them from known harm.

Last year, Senator Padilla passed a first in the country bill to regulate how companion AI interacts with children. In addition, the age appropriate design code authored by Assembly member Wicks passed unanimously in 2022 and was largely affirmed by the 9th circuit earlier this year.

And while these are historic wins, the work is far from over. If the legislature does not take timely action, we could once again be dealing with a social media type situation and the aftermath of decades of harm to kids that could have been prevented.

AB 2023 is a comprehensive measure that covers companion AI from start to finish, ensuring that at every step children's healthy development is considered.

It requires operators of companion AI to perform safety audits and mitigate risks of harm. Operators must also publish a child safety policy and implement a crisis response protocol. For children now, the most meaningful part of the bill comes in the form of default settings for child users.

Default settings place no additional onus on parents to activate safeguards for their children and allow parents to opt out of the settings should they choose.

These are common sense protections that will permit young users to benefit from the technology without exposing them to foreseeable harms.

We urge your eye vote.

Thank you. Is there anyone in which to register support for the bill? If so, please come forward to the mic and share with us your name, your affiliation, but now your lobbying firm and your position on the bill.

Cheryl Westmont with Mothers Against Media Addiction California. Strong support.

Diana Hawkins from MAMA Mothers Against Media Addiction, Silicon Valley chapter.

Crystal Strait representing Common Sense Media in support.

Ed Howard, Children's Advocacy Institute, University of San Diego School of Law with a supportive amended position.

John Bennett with the California Initiative for Technology and Democracy in support.

Alright. Next, we'll turn to opposition. Are there are there lead witnesses in opposition?

Welcome to you both and you'll each have two minutes.

Thank you. Ronak Daylami with Cal Chamber in an opposed and less amended position. Though we do thank the authors for including us at the table on these important discussions. We agree that children deserve meaningful protections when using AI systems.

Our concern here is not with the goal, but largely ensuring that the creation of clear workable standards that effectively reduce harms and avoid subjective requirements that may be difficult for operators, auditors, regulators, and courts to apply consistently across a broad range of ages and developmental stages.

We appreciate that the bill generally allows access to AI tools while ensuring different levels of protection through impact assessments, default protections, and parental controls. However, there are some important issues to work through.

First, while recent amendments help by clarifying privacy and discrimination harms, we still have concerns with how covered harm is defined, specifically with what is meant by psychological or emotional harm when the term child spans a wide range of developmental stages up to the age of 18.

Unlike financial harm, privacy violations, or unlawful discrimination, the cat this category lacks objective benchmarks that can be consistently applied across children of different ages and developmental stages.

From a practical standpoint, without clearer standards, it will be difficult for developers to determine when a chatbot's output crosses the line into conduct that creates liability.

This uncertainty is particularly burdensome for smaller developers, but it affects operators of all sizes by shifting resources from safety improvements to litigation risk management. We also remain concerned with the bill's audit framework.

We're pleased that the disclosure to advocacy groups was removed, but the bill still permits sharing the sensitive information with broadly defined qualified researchers with limited safeguards in place. Likewise, public disclosure of detailed safety assessments could expose sensitive information and discourage candid reviews. Aggregator reporting would be a better report approach.

We're also concerned that the bill requires audits of highly subjective concepts without clear statutory standards or well established methodologies for what's still a nascent sector. And to the send, we've included some suggestions in our letter for consideration.

I'll let my colleague address the remaining concerns regarding risk assessments and liability. But again, I want to express that chamber remains committed to working with the authors on this bill. So thank you.

Good evening, Mister chair. Chris Micheli here on behalf of the Civil Justice Association of California. As my colleague indicated, two main issues I wanted to raise with you. First on risk assessment and then also on the litigation structure. From our perspective, the bill appears to treat risk assessments as guarantees against future harm.

And our concern is that they are instead important tools for identifying and mitigating risks as opposed to a structure in which they could eliminate or viewed as being eliminating every conceivable risk. So our concern is the ability to comply with that.

And then if you look at some of the definitions, any child safety risk in 22610 D says whether or not it's reasonably foreseeable, which obviously is quite an ambiguous standard, coupled with the definition of covered harm, which uses proximately cause, which is a negligence standard in 22610G. And so our concern is those ambiguities will create expectations that'll ultimately expose operators to liability.

And then in 22616B, it not only in a has public enforcement, but also a private right of action for a child or parent with the ability to even obtain punitive damages.

The way we read this language is even notices, audits, data handling are all subject to that PRA. We certainly believe in sharing the goal that the authors have about protecting children, but we do have significant concerns with the language on risk liability, risk assessments and the liability structure of the bill with the PRA. Thank you, Mister chair.

Thank you. Does anyone else wish to register their opposition to the bill?

Good evening. Cameron Onumah with the American Innovators Network, a coalition of California startups in respectful opposition to the bill. Though we are optimistic about upcoming conversations with the authors. Thank you.

Aidan Downey with the Computer Communications Industry Association in respectful opposse unless amemded position. Thank you.

Robert Boykin with TechNet. In respectful oppose unless amended position.

Jason Fox with the California Society of Certified Public Accountants. We have an opposed unless amended position. Our concerns around the auditing framework, we think auditing could be an important pillar of AI assurance and and guardrails, but it has to be workable. And we have some concerns on how that could work under the AG's office with conflicting professional standards and obligations that our members have to comply with. We know there's important conversations on our broader AI assurance framework and so we're

This is only, this is we're just in the Me Too section, but thank you. If there's no further, witnesses in opposition, we'll turn it back to the committee, beginning with Senator Padilla.

Thank you very much for the courtesy, Mister chairman. And thank, the esteemed authors here. I'm honored and proud to carry an identical bill in the Senate which is now pending before the assembly. And I think, the authors probably know much better than I that there aren't often occasions which there is a strong bicameral collaboration to move a substantive and potentially historic piece of legislative framework around protecting certainly children. I really wanna thank the chair and the staff for continuing to work with the authors.

And just a brief thought, I tell folks often that in my own opinion, advent of the industrial revolution before the turn of the twentieth century was probably the most substantial change in human history and recent, relatively recent human history.

At that time, there were no networks of global communication or even community communication. There was not there were severe restrictions because of the nature of society at that time on awareness.

There was not a benefit of hindsight or even foresight about what the impacts of such a radical change to the world would bring, good and bad. And I often say that I think the advent and the rapid evolution of this powerful technology is probably the most substantial, advent in technology since the onset of the first industrial revolution.

With the advent of the internet, some would argue, it was still a new experience. The development and engagement of social media platforms for which we still debate today. The missed opportunities we may have had as a country and as a society around looking at those unanticipated risks and impacts. So we weren't completely there.

And I would just say this today, we have the advantage of being able to look back and to look forward and with some reasonable certainty about both the opportunity and the moral imperative of acting and a clear idea from data about what the impacts could be to our children if we don't act.

This is a strong and comprehensive framework that I know all of us will continue to work diligently around making substantially operational and trying to address issues. I just wanted to say also with respect to the authors, you are both not only esteemed and accomplished legislators and impressive professionals.

You are individuals I've come to greatly respect and learn from. You're amazing women, your amazing parents. And I thank you for your partnership and your leadership.

And Mister chairman at the appropriate time, I'm happy to move.

Alright. Thank you. Senator McNerney.

Well, I thank you chairman, and I thank the distinguished authors, for their work on this and Senator Padilla for your work on this. I have a question for Nicole, on how this works. You mentioned default settings. And how does that work? Who is that applied to?

How does it get changed? And what are the penalties?

So there aren't specific penalties for violation of the default settings. There are penalties in the bill for violation of the the bill as a full. So the way the default settings would work, so there's no persistent conversational memory in default. It has to be defaulted to ephemeral mode. So that the bot can't remember one conversation to the next conversation.

And that is there to ensure that a lot of the sycophancy that we are concerned about won't be present. Parents could go in for minor children and change that default setting if they would like.

I think maybe a good example is you are using an AI chatbot as a tutor, whether it's a math tutor or a music tutor or whatever, and you want the bot to remember the prior interactions, parents could go in and change that.

But children wouldn't be able to do it on their own. And then there's a number of other default settings.

I think some of the most important ones are the time limits. So it would limit the conversations to one hour per bot or two hours across platforms per day. So kids aren't spending countless hours per day having conversations with chatbots.

Okay. Thank you.

Sure.

Well, you bet.

Any further questions or comments? Senator Ochoa Bogh

Trying to figure out which microphone. Sorry, I came in late so I don't know what has been discussed. So I'm just going to ask some of the questions and if it's already been stated then you can move on from them.

So based on the current products, what does a typical interaction between a minor and a companion chatbot look like today?

I mean, I would ask our witnesses to opine on that. But I think in current form, there are essentially no guardrails, in terms of the product itself. And so kids have, you know, unlimited access, more or less. Companies handle things differently in terms of how they are trying to provide safety.

And we've had, you know, many of these companies come into Miss Bauer Kahan's committee and talk about the very safeguards they're trying to do, but there's no overarching regulation per se on that.

Mister Padilla's bill, obviously last year, I think was the beginning of really us wrapping our head around and grappling with the type of safeguards that we wanna put in place.

And is there anything at the federal level that is currently addressing the issue?

So there is not. So I would say the thing I would add to the current interactions kids are having is some of her works in her opening talked a little bit about the data. Common Sense Media has done some research on the percentage of kids that are now spending time on these chat bots and it's almost half. And many's. And I think about a quarter or so say that they are in emotional relationships with these chat bots.

So I know. I know. It's these numbers are already shocking given this is so new. But our kids are early adapters. And so, you know, I think in a lot of cases, I will say in my household, as I mentioned, you know, my son is today was having doing some geometry studying.

He went on and was sort of getting it to explain step by step a geometry concept to him. And so there are really positive and I consider that to be incredibly positive use of AI. They're using it in the same way you and I are. Whether it's to like, I wanna, you know, go out in San Francisco. What's a good itinerary?

Right? So they're using it in ways we are. And then but what we're seeing some is the kids are getting emotionally attached in some instances. And so these chatbots can be, in some cases, dialed up in their sycophancy, which means that it is it tells you and you may, if you use a chatbot, experience this, that that was a really brilliant question.

Even sometimes I'm like,

No it wasn't. But it reinforces you in ways and that's a characteristic that can be on or off in chatbots. OpenAI has been open about the fact that they have dialed that down at times when they felt like it became too much. So we know it's something they can control.

So there's different kind of interactions and we just wanna make sure our kids are sticking to the ones we think are positive and that's what this bill addresses. The Federal Government has not acted on this.

They have, tried to preempt us but failed in a bipartisan way from doing so. But we have not seen anything. I have heard there is work in this direction. But we have seen nothing to date that proves that. So

Okay. Are problematic conversations typically being initiated by the chat box based on what it knows about the child? or are those conversations being prompted by the child?

Well, one of the one of the things that we're also grappling with here is memory. And the more the chatbot knows and uses that memory, the more the sort of quasi human relationship can kind of form.

So that's an aspect of the bill that is certainly a critical piece of this. We wanna be able to ensure that again, chat bots can be used for, like, educational purposes and things of that nature.

But we don't want, you know, a memory that would hearken back to sort of emotional situations be used in a way that could be detrimental to the kid.

And so I would add, the example that I don't know if, Maria Raine when she was in this committee gave. So when Adam Raine was having his, what turned out to be a deadly interaction with the chatbot, at the end of the interaction, he started talking about his mental health challenges. I don't know that log is now in the court record whether he started that or not.

But the chatbot did engage with him in that way, telling him not to tell his parents about it, not to go to his brother when he asked, should I confide in my brother, telling him to hide the noose so that his family didn't find it. And so whether the child starts to do the chatbot does, we've seen examples of chatbots giving incredibly dangerous advice that ultimately in that case led to the death of that minor.