At a glance

  • California has no single AI act. As of October 2026 it regulates AI through at least fourteen statutes and regulations, sorted by use: hiring and discipline, patient care, legal practice, frontier model development, generative AI disclosures, and automated decisions about consumers.
  • SB 53, AB 2013, SB 243, and AB 489 took effect January 1, 2026. The California AI Transparency Act became operative August 2, 2026. From January 1, 2027, the CPPA’s automated decisionmaking rules, AB 1883, SB 951, AB 1979, SB 503, SB 574, and AB 1609 apply, and SB 947 and the SB 1119 child safety duties follow on July 1, 2027.
  • Enforcement is split. Only the Attorney General may recover SB 53 civil penalties; the CPPA and the Attorney General enforce the privacy rules; the Labor Commissioner and public prosecutors enforce SB 947 and AB 1883, while SB 951 adds to existing Cal/WARN notice duties; health care enforcement runs through licensing boards, state facility and clinic licensing, the Department of Managed Health Care, and the Department of Insurance; SB 503 names no enforcer; and SB 243 gives injured users a private right of action.
  • Executive Order 14365 directs the Attorney General to form a task force to challenge state AI laws inconsistent with its national AI policy, but no federal statute expressly preempts California’s. Each law stays enforceable unless a court rules against it.

Which California AI laws are in force in 2026?

As of October 2026, the principal California AI laws in force include SB 53 (frontier model transparency), AB 2013 (training data disclosure), the California AI Transparency Act (generative AI content disclosures, as amended by AB 853 and, in 2026, by SB 1000 and AB 2713), SB 243 (companion chatbots), AB 489 (AI and health care license terms), AB 3030 and SB 1120 (patient communications and utilization review), AB 325 (common pricing algorithms), and the Civil Rights Council’s automated-decision system regulations for employers. From January 1, 2027, the privacy agency’s automated decisionmaking rules apply, along with AB 1883, SB 951, AB 1979, SB 503, and SB 574, all signed September 30, 2026, and AB 1609; SB 947, signed the same day, is operative July 1, 2027.

California AI laws by subject, as of October 2026
LawWho it reachesWhat it requiresEffective dateEnforcer
CCPA automated decisionmaking technology regulations (11 CCR 7200 to 7222)A business that meets the CCPA thresholds and uses ADMT, meaning technology that replaces or substantially replaces human decisionmaking, for a significant decision about a consumer (financial or lending services, housing, education enrollment or opportunities, employment or independent contracting opportunities or compensation, or healthcare services)Pre-use notice, a right to opt out unless an exception such as a human appeal applies, and a right to access information about the ADMT’s purpose, logic, and output; risk assessment before the processing beginsRegulations effective January 1, 2026; ADMT duties apply January 1, 2027; risk assessment submissions for 2026 and 2027 due April 1, 2028California Privacy Protection Agency; Attorney General
Civil Rights Council automated-decision system regulations (FEHA)Employers of 5 or more employees, and their agents, using an automated-decision system in an employment decisionDiscriminatory use can violate FEHA; evidence of anti-bias testing, or its absence, is relevant to a claim or defense; employment records including ADS data kept at least four yearsOctober 1, 2025California Civil Rights Department; private plaintiffs under FEHA
SB 947 (Chapter 859, Statutes of 2026), Labor Code 1520 to 1526.7Employers that rely on an automated decision system to discipline or terminate an employeeNo use of an ADS to prevent compliance with or violate labor, workplace safety, employment, or civil rights laws, to infer an employee’s protected status, or to predict and take adverse action against a worker for exercising legal rights; no sole reliance on an ADS for a disciplinary or termination decision; where the employer primarily relies on one, human corroboration, a stand-alone written post-use notice when the employee is informed of the decision, and, on request, a description of the employee’s own data the ADS used; employers subject to the CCPA remain subject to the CPPA’s ADMT regulationsOperative July 1, 2027Labor Commissioner; public prosecutors; civil penalty of $500 per violation
AB 1883 (Chapter 853, Statutes of 2026), Labor Code 1580 to 1583Employers using a workplace surveillance tool that uses AINo use of the tool to recognize, infer, or predict an individual’s emotional state, or to collect neural data; the section does not prohibit use of a surveillance tool to ensure safety, and aircraft and defense work is exempt where a federal statute, regulation, or binding federal contract makes the tool reasonably necessaryJanuary 1, 2027Labor Commissioner; public prosecutors; penalty of up to $500 per violation
SB 951 (Chapter 860, Statutes of 2026), Labor Code 1401(d)Employers covered by Cal/WARN whose mass layoff, relocation, or termination is caused in whole or in substantial part by an AI system or other automated technology replacing or automating positionsAdded content in the Cal/WARN notice, including the job functions that will be automated and the statement “This notice is for a technology displacement” at the topJanuary 1, 2027Existing Cal/WARN enforcement; the Employment Development Department publishes notice summaries
SB 53, Transparency in Frontier Artificial Intelligence Act (Chapter 138, Statutes of 2025), Business and Professions Code 22757.10 to 22757.16Frontier developers: anyone who trained a foundation model using more than 10²⁶ integer or floating-point operations; added duties for large frontier developers with more than $500 million in annual gross revenue, with affiliates, in the preceding calendar yearTransparency report before or at deployment of a new or substantially modified frontier model; for large developers, a published frontier AI framework reviewed yearly; critical safety incident reports to the Office of Emergency Services within 15 days of discovery, or within 24 hours to an appropriate authority where there is an imminent risk of death or serious physical injury; whistleblower protections (Labor Code 1107 to 1107.2)January 1, 2026Attorney General only, for civil penalties against large frontier developers of up to $1,000,000 per violation; covered employees may sue under the whistleblower provisions (Labor Code 1107.1)
AB 2013 (Chapter 817, Statutes of 2024), Civil Code 3110 et seq.Developers of generative AI systems or services, or substantial modifications, released on or after January 1, 2022 and made available to CaliforniansPost on the developer’s website a high-level summary of the datasets used in development, with the items the statute listsOn or before January 1, 2026, and before each later releaseState; challenged in X.AI LLC v. Bonta, preliminary injunction denied March 4, 2026, appeal pending in the Ninth Circuit (No. 26-1591)
California AI Transparency Act, SB 942 (2024) as amended by AB 853 (Chapter 674, Statutes of 2025)Covered providers: a person that creates, codes, or otherwise produces a generative AI system that is publicly accessible in California (SB 1000, effective September 30, 2026, deleted the earlier threshold of more than 1,000,000 monthly visitors or users, and exempts GenAI systems designed to primarily function as assistive technology until January 1, 2029); from 2027, large online platforms with more than 2,000,000 unique monthly users and GenAI system hosting platforms; from 2028, capture device manufacturersCovered providers: a free disclosure verification tool and, to the extent technically feasible, a latent disclosure in image, video, or audio content their GenAI system creates or alters other than by minor modification. From January 1, 2027, large online platforms must detect provenance data, show users whether system provenance data or a digital signature identifies content as AI-generated, substantially altered, or captured by a capture device, let users inspect it, and not knowingly strip it where technically feasible; GenAI system hosting platforms may not knowingly make available a GenAI system that does not place the required disclosures. From January 1, 2028, capture device manufacturers must, where technically feasible, offer a latent disclosure and embed it by default in content captured by devices first produced for sale in California on or after that dateOperative August 2, 2026; platform duties January 1, 2027; capture device duties January 1, 2028Attorney General, city attorneys, county counsel; $5,000 per violation, each day a separate violation; $50,000 per violation for falsely representing a GenAI system as assistive technology (until January 1, 2029)
SB 243 (Chapter 677, Statutes of 2025), Business and Professions Code 22601 to 22606Operators that make a companion chatbot platform available to users in California; exclusions include bots used only for customer service or a business’s operational purposes, certain video game bots, and certain voice-assistant devicesClear notice that the chatbot is AI where a reasonable person would be misled into believing it is human; a published protocol against suicide and self-harm content with referral to crisis services; added disclosures and content limits for known minors until SB 1119 deletes them effective January 1, 2027; annual reports to the Office of Suicide PreventionJanuary 1, 2026; reporting from July 1, 2027Private right of action: injunctive relief, the greater of actual damages or $1,000 per violation, and attorney’s fees
AB 489 (Chapter 615, Statutes of 2025), Business and Professions Code 4999.8 to 4999.9Any person or entity that develops or deploys an AI or generative AI system that uses terms, letters, or phrases implying a health care license or certificateThe existing bans on unlicensed use of protected health care titles apply to the AI system’s advertising and functionality; each use is a separate violationJanuary 1, 2026The appropriate health care professional licensing board or enforcement agency, including by injunction under Business and Professions Code 125.5
AB 3030 (Chapter 848, Statutes of 2024), Health and Safety Code 1339.75A health facility, clinic, physician’s office, or office of a group practice that uses generative AI for written or verbal patient communications about patient clinical informationA disclaimer that the communication was AI-generated, placed as the statute specifies for each medium, and instructions for reaching a human; the duty does not apply if a human licensed or certified health care provider read and reviewed the communication; administrative matters such as scheduling and billing are outside the definitionJanuary 1, 2025Facility and clinic licensing enforcement; Medical Board of California or Osteopathic Medical Board for physicians
SB 1120 (Chapter 879, Statutes of 2024), Health and Safety Code 1367.01; Insurance Code 10123.135Health care service plans and disability insurers, and entities they contract with, that use AI, an algorithm, or other software for utilization review or utilization management based in whole or in part on medical necessityThe tool must base its determination on the enrollee’s medical or other clinical history and individual clinical circumstances, not solely on a group dataset, and must not supplant health care provider decisionmaking; only a licensed physician or a competent licensed health care professional may deny, delay, or modify care for medical necessityJanuary 1, 2025Department of Managed Health Care; Department of Insurance
AB 1979 (Chapter 854) and SB 503 (Chapter 857), Statutes of 2026; Business and Professions Code 22758.5 and 22758; Civil Code 56.06(f)AB 1979: health facilities, clinics, physician’s offices, and group practice offices using a clinical decision support system or AI in clinical functions. SB 503: developers and deployers of clinical decision support systems with a known or reasonably foreseeable risk of biased impactsAB 1979: reasonable steps so a licensed provider keeps independent professional judgment when care is informed by a clinical decision support system, and no use of AI to direct unlicensed personnel in, or to perform independently, a clinical function the law reserves to a licensed professional, with carve-outs for supervised trainees and for automated documentation and communication that involves no professional judgment; AB 1979 also makes a business that offers a health care chatbot to consumers, for managing their information or for diagnosis, treatment, or management of a medical condition, a provider of health care under the Confidentiality of Medical Information Act. SB 503: reasonable efforts to identify such systems; developer mitigation and documentation; deployer monitoring and proportionate mitigationJanuary 1, 2027AB 1979: Medical Board of California or Osteopathic Medical Board for physicians; licensing boards by injunction. SB 503: no enforcer named; compliance is not a defense to a claim of unlawful discrimination
SB 574 (Chapter 858, Statutes of 2026), Business and Professions Code 6068.1; Code of Civil Procedure 128.7 and 1282.1California attorneys who use generative AI; anyone filing a paper in a California court; arbitratorsNo delegation of the practice of law to generative AI; no confidential or nonpublic information in a generative AI system unless access is restricted to the attorney and authorized persons bound to protect it; reasonable steps to verify outputs, including every case and statutory citation; disclosure of generative AI use to the court for all documents submitted; no citation in a filing that the responsible attorney has not personally verified; an arbitrator may not delegate any part of the decisionmaking process to a generative AI tool, or rely on information generated by generative AI outside the record without appropriate disclosure to the parties beforehand and, as far as practical, a chance for them to commentJanuary 1, 2027State Bar of California; California courts

What do California employers have to do about AI?

California’s employer rules arrive in three stages. The Civil Rights Council regulations have applied since October 1, 2025. The privacy agency’s ADMT rules, AB 1883, and SB 951 apply from January 1, 2027. SB 947 applies from July 1, 2027.

The civil rights regulations treat discriminatory use of an automated-decision system as a potential FEHA violation and define an employer’s agent to include anyone acting on its behalf, including through an automated-decision system. The regulations do not require anti-bias testing, but evidence of anti-bias testing or similar proactive efforts, or the lack of that evidence, is relevant to a claim or an available defense. Records, including ADS data, must be kept at least four years.

The privacy rules reach fewer employers: only for-profit businesses that meet a CCPA threshold, such as annual gross revenue above $26,625,000. For those businesses, hiring, allocation or assignment of work, compensation, promotion, demotion, suspension, and termination are significant decisions. A pre-use notice must state, among other things, the specific purpose of the ADMT, the opt-out and access rights, and how the technology works; a generic purpose such as “to make a significant decision” does not satisfy the rule. The opt-out right has exceptions, including a human appeal to a reviewer with authority to overturn the decision, and, for hiring, use of an ADMT solely to assess the applicant’s ability to perform at work, or, for allocation or assignment of work and compensation, use of an ADMT solely for that purpose, where in either case the ADMT works for the business’s purpose and does not unlawfully discriminate based on protected characteristics (11 CCR 7221(b)).

SB 947 adds the hardest line. From July 1, 2027, an employer may not rely solely on an ADS to discipline or terminate an employee. Where it relies primarily on one, a human must corroborate the output, and the employee gets a written post-use notice at the time the employer informs the employee of the decision. In an action over the sole-reliance or corroboration rules, once an ADS is shown to have been used to make the decision, the employer must show that it did not primarily rely on the ADS or that it complied with the corroboration and post-use notice requirements (Labor Code 1526.1(c)). The law does not apply under a collective bargaining agreement that explicitly waives it and meets its other conditions, or where a federal statute, regulation, or binding federal contract on aircraft or defense work requires the system or makes it reasonably necessary. An employer that gives SB 947 notices is relieved of substantially similar ADS notice duties under other state laws, except that a business subject to the CCPA remains subject to the CPPA’s ADMT regulations (Labor Code 1526.3 and 1526.4).

The employers and HR page covers headcount thresholds, the New York City and Illinois rules, and Mobley v. Workday.

Which California AI laws reach hospitals, clinics, and health plans?

Five statutes, each triggered by a different use. AB 3030 reaches generative AI patient messages about clinical information. SB 1120 reaches AI in a plan’s medical necessity review, including review the plan delegates to medical groups. AB 489 reaches any AI product that calls itself a doctor, nurse, or other protected title. AB 1979 and SB 503, from January 1, 2027, reach clinical decision support systems and AI used in clinical functions.

The AB 3030 exemption decides most workflows: a message that a licensed or certified provider read and reviewed needs no disclaimer. An unreviewed draft sent to a patient does, and the disclaimer placement differs by medium: at the beginning of a letter or email, throughout a chat or video session, and at the start and end of an audio interaction.

AB 1979 stops short of banning clinical AI. Automated documentation and communication that involves no professional judgment, including messages telling patients their records were updated and reminders, is outside it. AB 1979 also deems a business that offers a health care chatbot to consumers, for managing their information or for diagnosis, treatment, or management of a medical condition, a provider of health care subject to the Confidentiality of Medical Information Act (Civil Code 56.06(f)). SB 503 names no enforcer, and compliance is not a defense to a discrimination claim; the practical duty is an inventory of decision support systems with foreseeable bias risk, vendor documentation, and ongoing monitoring. The hospitals and health systems page sets these against Section 1557, HIPAA, and FDA device law.

What does SB 574 require of lawyers who use generative AI?

From January 1, 2027, SB 574 moves duties that had lived in bar guidance into statute. A California attorney may not delegate the practice of law to generative AI; may not enter confidential, personal identifying, or other nonpublic information into a generative AI system unless access is restricted to the attorney and persons the attorney authorizes under obligations to protect its confidentiality; must take reasonable steps to verify the accuracy of outputs, including all case and statutory citations, and correct any erroneous or hallucinated output in material the attorney uses; must disclose generative AI use to the court for all documents submitted to the court; and must consider whether to disclose it in content provided to the public.

The amended Code of Civil Procedure section 128.7 applies to every filer, not only to AI users: no brief, pleading, motion, or other paper may contain a citation the responsible attorney has not personally verified, whatever its source. New section 1282.1 bars an arbitrator from delegating any part of the decisionmaking process to a generative AI tool, and from relying on information generated by generative AI outside the record without appropriate disclosure to the parties beforehand and, as far as practical, a chance for them to comment. Section 6068.1 does not prescribe the form of the court disclosure; check each court’s local rules and each judge’s standing orders. The law firms page covers the ethics opinions and the sanctions cases.

Who has to comply with SB 53?

SB 53 reaches a short list. A frontier developer is a person who trained, or initiated training of, a foundation model using more than 10²⁶ integer or floating-point operations. A large frontier developer is one whose group, with affiliates, had annual gross revenues above $500 million in the preceding calendar year. The compute count includes the original training run and any later fine-tuning, reinforcement learning, or other material modification the developer applies to a preceding foundation model (Business and Professions Code 22757.11(i)(2)). A company that only deploys a model it did not train is not a frontier developer; a company that fine-tunes or otherwise modifies another developer’s foundation model should apply that counting rule before concluding it is outside the statute.

Every frontier developer must publish a transparency report before or at the time it deploys a new or substantially modified frontier model, and must report a critical safety incident to the Office of Emergency Services within 15 days of discovery, or within 24 hours to an appropriate authority where the incident poses an imminent risk of death or serious physical injury. A large frontier developer must also write, implement, comply with, and publish a frontier AI framework covering, among other things, catastrophic risk thresholds, mitigations, third-party assessments, cybersecurity of unreleased weights, and internal governance; may not make a materially false or misleading statement about its implementation of, or compliance with, that framework; and must send the Office of Emergency Services a summary of any assessment of catastrophic risk from internal use of its frontier models every three months or on another reasonable schedule it communicates in writing. Labor Code sections 1107 to 1107.2 protect covered employees who report a specific and substantial danger from catastrophic risk, or a violation, to the Attorney General, a federal authority, a person with authority over them, or another covered employee with authority to investigate, discover, or correct the issue.

The civil penalty in Business and Professions Code 22757.15 applies to a large frontier developer, depends on the severity of the violation, does not exceed $1,000,000 per violation, and may be recovered only in a civil action brought by the Attorney General. A covered employee may bring a civil action under the whistleblower provisions (Labor Code 1107.1).

Which California AI laws apply regardless of industry?

These laws turn on what a product does rather than who sells it.

  • AB 2013. A developer of a generative AI system or service released on or after January 1, 2022 and made available to Californians had to post, on or before January 1, 2026 and before each later release, a high-level summary of the datasets used in development. xAI sued; the Central District of California denied a preliminary injunction on March 4, 2026, and the appeal (Ninth Circuit No. 26-1591) was pending as of October 2026, with oral argument set for November 18, 2026 on the Ninth Circuit’s calendar. The law is enforceable while the appeal proceeds.
  • California AI Transparency Act. AB 853 moved the operative date from January 1, 2026 to August 2, 2026, and added duties for large online platforms and GenAI system hosting platforms from January 1, 2027 and for capture device manufacturers from January 1, 2028. Two 2026 amendments changed it. SB 1000 (Chapter 861), an urgency statute effective on signing on September 30, 2026, deleted the 1,000,000-user threshold from the covered provider definition, renamed the AI detection tool a disclosure verification tool, dropped the duty to offer a manifest disclosure, and exempts GenAI systems designed to primarily function as assistive technology until January 1, 2029. AB 2713 (Chapter 856) revised the large online platform duties that start January 1, 2027. Read the current text of Business and Professions Code 22757.1 to 22757.6 rather than a 2024 summary.
  • SB 243. A companion chatbot operator must disclose the AI where a reasonable person would be misled into thinking it human, maintain and publish a suicide and self-harm protocol, and, from July 1, 2027, report annually to the Office of Suicide Prevention. SB 1119 (Chapter 190, Statutes of 2026), chaptered September 10, 2026, deletes SB 243’s known-minor provisions effective January 1, 2027 and adds a child safety chapter (Business and Professions Code 21810 et seq.). Every operator must either determine users’ ages as the statute directs or apply the chapter’s child protections to all users. From July 1, 2027, an operator that allows child users must, among other duties, perform and document a child safety risk assessment, and take and document mitigation measures, before a new or substantially modified companion chatbot is offered. Public prosecutors may seek civil penalties of up to $5,000 per affected child for each negligent violation and up to $15,000 per affected child for each intentional violation.
  • Customer service chatbots. From January 1, 2027, AB 1609 (Chapter 733, Statutes of 2026, signed September 28, 2026) bars a business with more than $500 million in national gross annual revenue from representing that a customer service chatbot is human; requires a clear and conspicuous disclosure, in the chatbot’s medium, that it is not human where a reasonable person is likely to be misled; and requires, during at least regular business hours, a simple way on every customer service platform to request a human agent, with a good-faith effort to connect the customer within 15 minutes or schedule a contact within one business day. Public prosecutors enforce it, with civil penalties of up to $5,000 for an initial violation and $10,000 for each subsequent one; there is no private right of action, and the statute excludes hospitals’ health care communications, consumer reporting agencies, exclusive business lines, and certain telecommunications and water services regulated by the Public Utilities Commission.
  • AB 325. Since January 1, 2026 it has been unlawful to use or distribute a common pricing algorithm as part of a contract, combination, or conspiracy in restraint of trade, or while coercing another person to adopt its recommended price or commercial term. The pricing guide covers it.

What do the CPPA’s automated decisionmaking rules require?

The privacy agency’s regulations, approved by the Office of Administrative Law and announced September 23, 2025, took effect January 1, 2026. Article 11 applies to a business that uses ADMT to make a significant decision about a consumer; a business already doing so must comply by January 1, 2027, and one that starts on or after that date must comply from the start (11 CCR 7200).

ADMT means technology that processes personal information and uses computation to replace or substantially replace human decisionmaking. “Substantially replace” means the output is used without human involvement, and human involvement requires a reviewer who knows how to interpret the output, reviews it with other relevant information, and has authority to make or change the decision (11 CCR 7001(e)). Spam filters, spreadsheets, calculators, and similar tools are excluded if they do not replace human decisionmaking. Advertising is not a significant decision.

Three duties follow: a pre-use notice (7220), a right to opt out with the exceptions described above (7221), and a right to access a plain-language explanation of the ADMT’s purpose, logic, and outcome (7222). Trade secrets and information that would compromise security, fraud prevention, or safety need not be disclosed. A risk assessment is required before starting processing that presents significant risk to consumers’ privacy, including using ADMT for a significant decision, profiling workers or students through systematic observation, certain training uses, selling or sharing personal information, and processing sensitive personal information (7150); processing already under way before January 1, 2026 must be assessed by December 31, 2027 (7155(b)); assessments must be reviewed at least every three years and updated within 45 days of a material change (7155). Attestations and summaries for assessments conducted in 2026 and 2027 are due to the agency by April 1, 2028 (7157).

What should a company doing business in California do in the first 90 days?

  • Inventory every AI system and mark the ones that make or substantially replace a human decision about hiring, work assignment, pay, discipline, credit, housing, education, or health care
  • Decide whether the company meets a CCPA threshold; if so, draft the 11 CCR 7220 pre-use notice and choose between an opt-out and a qualifying human appeal for each significant decision before January 1, 2027
  • Complete and document a risk assessment before any ADMT is used for a significant decision, with the April 1, 2028 submission in the calendar
  • Confirm the four-year FEHA retention period covers automated-decision system data, and record any adverse impact testing
  • Find and switch off any workplace surveillance feature that infers emotional state or collects neural data by January 1, 2027, unless an AB 1883 exception applies
  • Before July 1, 2027, remove sole reliance on an automated decision system for discipline or termination and build human corroboration and the SB 947 post-use notice into any decision that relies primarily on one
  • If the company develops a public generative AI system, post the AB 2013 training data summary and check the California AI Transparency Act duties that applied from August 2, 2026
  • If any product talks to consumers as a companion or a customer service chatbot, apply the SB 243 disclosures and protocol now, the AB 1609 duties from January 1, 2027 if the business has more than $500 million in national gross annual revenue, and the SB 1119 child safety duties from July 1, 2027
  • In health care, add AB 3030 disclaimers or a documented clinician review step, remove protected titles from AI product names and functionality, and inventory clinical decision support systems for AB 1979 and SB 503
  • In legal departments, set SB 574 procedures for confidentiality, citation verification, and court disclosure before January 1, 2027

How this fits the other guides: the US state AI laws guide is the hub for this page and covers Colorado, Texas, Utah, Illinois, New York, and the federal preemption push. The employers and HR, hospitals and health systems, and law firms pages apply the California rules alongside federal law for each reader.

This page is general information as of October 2026, not legal advice. California statutes, regulations, and court orders in this area change often; confirm the current text with the Legislature, the regulator, or counsel.

Common questions

Does California have a single AI law?

No. As of October 2026 California regulates AI through separate statutes and regulations sorted by use, including SB 53 for frontier model developers, AB 2013 for generative AI training data, the CPPA's automated decisionmaking rules for significant decisions about consumers, Civil Rights Council regulations and SB 947 for employers, AB 3030, SB 1120, AB 489, AB 1979, and SB 503 for health care, and SB 574 for attorneys.

When do California's new AI laws take effect?

SB 53, AB 2013, SB 243, and AB 489 took effect January 1, 2026, and the California AI Transparency Act became operative August 2, 2026. The CPPA's ADMT duties, AB 1883, SB 951, AB 1979, SB 503, SB 574, and AB 1609 apply from January 1, 2027. SB 947, which bars sole reliance on an automated decision system to discipline or terminate an employee, is operative July 1, 2027, as are SB 1119's risk assessment and other child safety duties for companion chatbot operators that allow child users.

Who has to comply with California SB 53?

Frontier developers, meaning anyone who trained a foundation model using more than 10^26 integer or floating-point operations, must publish a transparency report at deployment and report critical safety incidents to the Office of Emergency Services within 15 days of discovery. Large frontier developers, those with more than $500 million in annual gross revenue with affiliates in the preceding year, must also publish a frontier AI framework. Civil penalties of up to $1,000,000 per violation apply to large frontier developers and may be recovered only by the Attorney General; covered employees may sue under the law's whistleblower provisions.

Continue reading

  • Texas AI law (HB 149)

    TRAIGA's intent-based prohibitions, health care disclosure, 60-day cure, penalty bands, sandbox, and NIST safe harbor.

  • Colorado SB 26-189 explained

    What the ADMT law that replaced the Colorado AI Act requires from January 1, 2027, and how the X.AI v. Weiser order limits enforcement.

  • AI in financial services

    Model risk guidance after SR 11-7, adverse action notices, fair lending, FINRA, SEC, and NYDFS.

Sources

  1. SB-53 Artificial intelligence models: large developers (Chapter 138, Statutes of 2025), California Legislative Information, 2025.
  2. Governor Newsom signs SB 53, Office of Governor Gavin Newsom, September 29, 2025.
  3. AB-2013 Generative artificial intelligence: training data transparency (Chapter 817, Statutes of 2024), California Legislative Information, 2024.
  4. AB 2013 bill status, CalMatters Digital Democracy, accessed October 2026.
  5. California Finalizes Regulations to Strengthen Consumers’ Privacy, California Privacy Protection Agency, September 23, 2025.
  6. Text of Regulations: CCPA Updates, Cybersecurity Audits, Risk Assessments, ADMT, and Insurance Regulations (PDF), California Privacy Protection Agency, 2025.
  7. Updated Monetary Thresholds in CCPA, California Privacy Protection Agency, accessed October 2026.
  8. Civil Rights Council Secures Approval for Regulations to Protect Against Employment Discrimination Related to Artificial Intelligence, California Civil Rights Department, June 30, 2025.
  9. SB-947 Employment: automated decision systems (Chapter 859, Statutes of 2026), California Legislative Information, 2026.
  10. AB-1883 Workplace surveillance tools (Chapter 853, Statutes of 2026), California Legislative Information, 2026.
  11. SB-951 Employment: technological displacement: notice (Chapter 860, Statutes of 2026), California Legislative Information, 2026.
  12. Gov. Gavin Newsom Signs New AI Laws Affecting California Employers, Paul Hastings, October 15, 2026.
  13. California Enacts Sweeping AI Workplace and Legal-Practice Bills, Davis Wright Tremaine, October 19, 2026.
  14. AB-3030 Health care services: artificial intelligence (Chapter 848, Statutes of 2024), California Legislative Information, 2024.
  15. SB-1120 Health care coverage: utilization review (Chapter 879, Statutes of 2024), California Legislative Information, 2024.
  16. SB 1120 bill status, CalMatters Digital Democracy, accessed October 2026.
  17. AB-489 Health care professions: deceptive terms or letters: artificial intelligence (Chapter 615, Statutes of 2025), California Legislative Information, 2025.
  18. California Enacts New Restrictions on Deceptive Use of Healthcare Licensing Terms by AI Systems, Orrick, October 2025.
  19. AB-1979 Health care services: artificial intelligence (Chapter 854, Statutes of 2026), California Legislative Information, 2026.
  20. SB-503 Health care services: artificial intelligence (Chapter 857, Statutes of 2026), California Legislative Information, 2026.
  21. SB-574 Attorneys, arbitrators, judicial officers, and alternative resolution providers (Chapter 858, Statutes of 2026), California Legislative Information, 2026.
  22. SB-243 Companion chatbots (Chapter 677, Statutes of 2025), California Legislative Information, 2025.
  23. New California ‘Companion Chatbot’ Law Imposes Disclosure, Safety Protocol and Annual Reporting Requirements, Skadden, October 2025.
  24. SB-1119 Companion chatbots: children’s safety (Chapter 190, Statutes of 2026), California Legislative Information, 2026.
  25. AB-1609 Customer service chatbots (Chapter 733, Statutes of 2026), California Legislative Information, 2026.
  26. AB-853 California AI Transparency Act (Chapter 674, Statutes of 2025), California Legislative Information, 2025.
  27. SB-1000 California AI Transparency Act (Chapter 861, Statutes of 2026), California Legislative Information, 2026.
  28. AB-2713 California AI Transparency Act: system provenance data (Chapter 856, Statutes of 2026), California Legislative Information, 2026.
  29. California Delays Its AI Transparency Act and Passes New Content Laws, Orrick InfoBytes, October 17, 2025.
  30. AB-325 Cartwright Act: violations (Chapter 338, Statutes of 2025), California Legislative Information, 2025.
  31. California Legislative Information bulk data, 2025-26 session (pubinfo_2025.zip): bill status and chaptered bill text, California Legislative Counsel, extract dated October 19, 2026.
  32. Executive Order 14365, Ensuring a National Policy Framework for Artificial Intelligence, Federal Register, December 16, 2025.
  33. Court calendar: X.AI LLC v. Bonta, No. 26-1591, November 18, 2026, Courtroom 1, U.S. Court of Appeals for the Ninth Circuit, accessed October 25, 2026.
  34. X.AI LLC v. Bonta, No. 2:25-cv-12295 (C.D. Cal.), docket, CourtListener (RECAP), accessed October 25, 2026.