Chatbots, Data Centers and Surveillance: 5 Silicon Valley Bills Land on Newsom’s Desk

California lawmakers approved a sweeping set of AI chatbot safety requirements for children and teens, known as Adam's Law (SB 1119), sending it to Gov. Gavin Newsom, who has until Sept. 30 to sign or veto. The bill, co-authored by Sen. Steve Padilla and Assemblymembers Rebecca Bauer-Kahan and Buffy Wicks, mandates age checks, safety risk assessments, independent audits, and default protections for young users, and is named after Adam Raine, a California teenager who died by suicide in 2025. OpenAI supports the bill, while the California Chamber of Commerce, Software & Information Industry Association, and TechNet oppose it.

Chatbots, Data Centers and Surveillance: 5 Silicon Valley Bills Land on Newsom’s Desk California’s Legislative session /politics ended at midnight on Monday, and numerous bills ranging from housing to healthcare to technology cleared both the Senate and the Assembly. Gov. Gavin Newsom has until Sept. 30 to decide whether he’ll veto or sign the measures. Here’s where some of the most closely watched tech /news/tag/technology bills landed. Companion chatbots and children’s safety California lawmakers on Monday approved a sweeping set of safety requirements for artificial intelligence chatbots used by children and teens. The bill, known as Adam’s Law https://calmatters.digitaldemocracy.org/bills/ca 202520260sb1119 , builds on last year’s SB 243 /news/12052617/newsoms-tightrope-walk-between-ai-regulation-and-silicon-valley-cash , which required chatbot operators to disclose when users were interacting with AI companions and to maintain protocols for detecting self-harm content. Co-authored by San Diego Democrat Sen. Steve Padilla and Assemblymembers Rebecca Bauer-Kahan D-Orinda and Buffy Wicks D-Oakland , the bill adds requirements to check users’ ages, safety risk assessments before new or substantially modified companion chatbots are released, independent audits and crisis-support resources. The bill also requires default protections for young users, including limits on notifications, usage and persistent conversational memory. It restricts targeted advertising to children and imposes additional privacy protections. Chatbot operators could face liability for failing to take reasonable measures to prevent certain harmful outputs, including content involving self-harm, sexually explicit material, romantic role-playing and emotionally manipulative behavior. Families would have a private right of redress for certain violations. The legislation is named for Adam Raine /news/12080610/california-mom-who-lost-her-son-to-an-ai-chatbot-is-now-fighting-to-regulate-them , a California teenager who died by suicide in 2025. His parents have said that Adam spent months discussing his loneliness and plans to harm himself with ChatGPT before his death. In a joint statement released after the bill successfully passed through the Legislature, Maria Raine called it “a serious and much-needed step forward for AI companion chatbot safety.” This year, lawmakers also passed another bill by Padilla that enacts a four-year moratorium on the sale and manufacturing of AI-chatbot-powered toys over concerns that the technology can harm children. The technology industry has not been uniformly opposed. OpenAI supports SB 1119 and has urged Newsom to sign it, saying in a statement the legislation establishes meaningful safeguards while preserving young people’s ability to use AI for learning, creativity and other purposes. “Teens are the first generation growing up with AI. Nearly nine in ten teens who use ChatGPT turn to it for learning, information, skill-building, or productivity in a given week. That opportunity comes with a clear responsibility: teens should have experiences designed around their distinct developmental needs, with strong default protections and appropriate opportunities to learn, create, and explore,” the statement said. Other technology and business groups, including the California Chamber of Commerce, Software & Information Industry Association and TechNet, have opposed the legislation. Newsom vetoed a stronger child-safety bill /news/12059714/newsom-vetoes-most-watched-childrens-ai-bill-signs-16-others-targeting-tech last year after raising concerns that its restrictions could effectively prevent teenagers from accessing conversational AI. Though his office has not indicated whether Newsom will sign, the version now on his desk was negotiated with those concerns in mind. Data centers and ratepayer protections California lawmakers on Monday approved two linked bills that would require the state to establish special rules for how large data centers pay for electricity and the infrastructure needed to serve them. SB 886 https://calmatters.digitaldemocracy.org/bills/ca 202520260sb886 , also authored by Padilla, and AB 2383 https://calmatters.digitaldemocracy.org/bills/ca 202520260ab2383 , by Assemblymember Rick Chavez Zbur D-Los Angeles , would require the California Public Utilities Commission to create separate electricity rates and updated interconnection rules for data centers. The measures are designed to prevent the costs of new power generation and grid upgrades needed to serve the facilities from being shifted onto other electricity customers. AB 2383 is contingent on SB 886 becoming law. “Big Tech keeps promising to be good neighbors, and these bills will make those promises legally enforceable,” Padilla said in a press release celebrating the passage of SB 886 and a related bill, 887 https://calmatters.digitaldemocracy.org/bills/ca 202520260sb887 . “No more saying one thing in the White House and doing the opposite in our local communities.” The legislation follows a year of intensifying political pressure over the expansion of data centers, which are being built to meet the enormous computing demands of AI. California has fewer giant data centers than Virginia and Texas, but the state still has one of the largest counts overall /forum/2010101912379/californias-ai-data-centers-taking-growing-environmental-toll . The facilities have generated growing concern /news/12028158/crackdown-on-power-guzzling-data-centers-may-soon-come-online-in-california over their electricity and water use, as well as the lack of transparency /news/12064265/new-report-reveals-strain-of-data-centers-on-californias-grid about their potential effect on utility bills. A Public Policy Institute of California poll released in July found that 73% https://www.ppic.org/publication/ppic-statewide-survey-californians-and-the-environment-july-2026/ of Californians opposed building data centers in their communities. In March 2026, the Little Hoover Commission released Data Centers and California Electricity Policy https://lhc.ca.gov/wp-content/uploads/LHC-Report-292-Data-Centers-California-Electricity-Policy-FINAL-PUBLIC-3.3.26.pdf , a report that explicitly recommends making tech companies and data center developers cover their own grid integration costs rather than shifting the burden to everyday ratepayers. “California’s electricity rates are among the highest in the nation, and affordability remains a top concern for residents,” the report states. Supporters — including The Utility Reform Network, Union of Concerned Scientists, Natural Resources Defense Council and utility worker unions — argued that companies building the facilities should pay for the infrastructure their electricity demand requires. Bill sponsor TURN pointed to a California grid-planning estimate that increased demand from data centers in PG&E’s service territory, where most of the state’s projects are concentrated, could generate as much as $1.8 billion in transmission-upgrade costs. The Data Center Coalition, which represents companies including Google and Microsoft, opposed the legislation, arguing that it unfairly singled out one type of electricity customer. PG&E also opposed the original approach, warning that it could increase costs for customers and delay infrastructure needed to meet the state’s broader energy demand. Business groups argued that California’s high energy costs, limited land and additional regulation could make the state less attractive for data-center development. The bills were substantially worked over in the final weeks of the session. Earlier drafts were far more prescriptive mandatory on-site batteries, 100% carbon-free power for data centers by 2030 , provisions that were stripped out in favor of giving the CPUC until July 1, 2027, to establish the special tariffs and rules. That shift gave Newsom and regulators more control over how the new rate structure will work. The final compromise drew criticism from some environmental and consumer advocates concerned the changes could open up loopholes that allow data centers to avoid paying some grid costs and wriggle out of clean-energy requirements for facilities that generate their own electricity. Newsom’s position has shifted noticeably during the course of the debate. Last year, he vetoed https://www.gov.ca.gov/wp-content/uploads/2025/10/AB-93-Veto.pdf legislation requiring data centers to disclose their water consumption, writing: “I am reluctant to impose rigid reporting requirements about operational details on this sector without understanding the full impact on businesses and the consumers of their technology.” This year, however, Newsom became directly involved in negotiations over both bills. On Monday in the state capitol, he said his administration had been working with lawmakers on several data-center bills and that the final versions “landed fairly well,” Politico reported https://www.politico.com/news/2026/09/01/california-lawmakers-big-swing-newsom-data-centers-01058848 . “You’re seeing states all across this country leaning in,” Newsom said. “We’ll be leaning in.” He also said last month that companies, rather than other electricity customers, should bear the costs associated with their data-center investments. “We believe that all of these investments, the costs, should be borne by the companies,” Newsom said. That makes a Newsom signature more likely here than it appeared earlier in the year, although the governor has not formally committed to signing either bill. Social media’s addictive features The California Senate voted https://calmatters.digitaldemocracy.org/bills/ca 202520260ab1709 to bar social media platforms like Instagram and TikTok from offering certain addictive features to users under 16. The measure, authored by Assemblymember Josh Lowenthal D-Long Beach , targets features on so-called “covered platforms” — which predominantly share user-generated content — designed to keep people engaged, including algorithmically-driven feeds and other mechanisms that can encourage prolonged use. It also would establish an e-Safety Advisory Commission within the California Department of Justice to advise state officials on online safety and report annually to the governor and Legislature. Common Sense Media, which sponsored the bill, is among its major supporters, along with Mothers Against Media Addiction and the American Academy of Pediatrics. The bill was amended as it moved through the Legislature, narrowing some of its provisions and changing how the state would oversee online safety. For supporters, the changes don’t go nearly far enough. Jim Steyer, founder and CEO of Common Sense Media, wrote Monday https://www.commonsensemedia.org/press-releases/statement-on-california-senate-judiciary-passage-of-ab-1709-to-protect-children-under-16-from-addictive that recent litigation /news/12096647/why-metas-landmark-settlement-could-be-social-medias-big-tobacco-moment against major technology companies has exposed how deliberately social media platforms have been designed to keep children engaged. “Social media companies have engineered their products to hook children, and their own researchers proved it worked,” Steyer said in a statement. “Parents cannot out-parent an algorithm built in a lab to keep kids scrolling.” The Electronic Frontier Foundation, Civil Justice Association of California and trade group NetChoice, which represents major technology companies including Google, Meta and Snap, oppose the measure. Civil liberties advocates, in particular, have raised First Amendment and privacy concerns, particularly around how platforms would determine which users are under 16 and which features should be considered addictive. “This well-intentioned, but deeply flawed piece of legislation, cuts young people off from essential information and experiences, particularly harming vulnerable youth and marginalized groups who often find safety in supportive online communities they can’t access offline,” the EFF wrote https://www.eff.org/deeplinks/2026/08/eff-gov-newsom-veto-californias-ab-1709 in a statement. Newsom has already signed several California laws aimed at protecting children online /news/12005803/newsom-signs-law-to-protect-children-from-social-media-addiction , including legislation requiring warning labels on social media platforms. But AB-1709 could face the same free speech and age-verification challenges that have complicated the state’s broader effort to regulate social media. A federal judge has already blocked portions of California’s existing law restricting addictive social media feeds while a legal challenge proceeds. Newsom has not publicly indicated whether he will sign the bill, but he called on lawmakers to explore stricter youth online safety guardrails during his final State of the State /news/12069094/in-final-state-of-state-speech-gov-newsom-says-california-offers-model-for-the-nation address. Workplace surveillance California employers could face new limits on the use of AI to monitor workers. AB 1883 https://calmatters.digitaldemocracy.org/bills/ca 202520260ab1883 , authored by Assemblymember Isaac Bryan D-Los Angeles , would prohibit employers from using AI-powered workplace surveillance tools to collect neural data https://law.stanford.edu/2024/12/02/what-are-neural-data-an-invitation-to-flexible-regulatory-implementation/ — sensitive personal information from a user’s nervous system — or recognize an employee’s emotional state, with some exceptions. It also covers certain other AI-based inferences about workers and allows civil penalties of up to $500 per violation. The measure is part of a small but growing effort by states to regulate how employers use AI to monitor and manage their workers. California’s bill is unusually focused on technologies that attempt to infer what is happening inside an employee’s body or mind. Other states have taken narrower approaches, including Illinois, which regulates AI-analyzed video interviews in hiring, and Maryland, which restricts the use of facial recognition technology. AB 1883 was sponsored by major state labor unions and Privacy Rights Clearinghouse. Business groups representing various employers opposed the bill, ranging from the American Petroleum and Convenience Store Association to the Wine Institute. The legislation’s ambitions were softened in committees. The latest version exempts certain workplace surveillance uses required by federal law, regulation or binding federal contracts involving aircraft, national security, military, space or defense work. Senate amendments also modified the bill’s enforcement provisions. Still, the core prohibition survived. “Workers should have a general expectation of privacy in certain areas,” said Lorena Gonzalez, who heads up the powerful California Federation of Labor Unions, to KQED’s Political Breakdown /news/12097046/california-labor-leader-on-ai-bills-and-billionaire-tax-measure-splitting-unions last week. Employers, Gonzalez said, are able to take advantage of outdated laws that don’t account for changes in technology. Labor interests have been at the forefront of legislation addressing AI anxieties /news/12076608/what-do-san-franciscos-ai-vs-humans-billboards-say-about-our-working-futures this session. “You don’t want the boss reading your mind,” Gonzalez added. The measure comes a year after Newsom vetoed the “No Robo Bosses Act,” which would have required employers to notify workers when AI was used in disciplinary or termination decisions and let employees challenge certain AI-generated decisions. In his veto message https://www.gov.ca.gov/wp-content/uploads/2025/10/SB-7-Veto.pdf last year, Newsom wrote the bill “imposes unfocused notification requirements on any business using even the most innocuous tools.” AB 1883 is narrower in scope. Rather than governing how AI makes decisions about workers, it bans a single, especially invasive use. Newsom has not said whether he will sign it. With a widely expected presidential run ahead, his decision will signal whether he’s willing to side with workers over employers and draw a hard line against AI that tries to read a worker’s mind.