Skip to main content
Gordon Rees Scully Mansukhani, LLP. logo.

California’s Expanding AI and Privacy Laws: Key Compliance Requirements for Businesses in 2027

California continues to expand its regulation of artificial intelligence (AI), automated decision-making technologies, and consumer privacy. In 2026, Governor Gavin Newsom signed a series of laws addressing AI-powered employment decisions, children’s online safety, companion chatbots, workplace surveillance, and the use of personal information in emerging technologies. Together with previously enacted legislation and regulations under the California Consumer Privacy Act (CCPA), these developments establish significant new compliance obligations for businesses beginning in 2027.

The evolving framework reflects a broader shift toward regulating how AI systems collect and process personal information, influence consequential decisions, and interact with consumers. Rather than imposing a single comprehensive AI law, California has adopted requirements that apply to different technologies, industries, and business activities.

Businesses should evaluate whether their existing privacy and AI governance programs adequately address these requirements, particularly where automated technologies are used in employment, consumer services, advertising, education, or products accessible to children.

Compliance Summary Table

Law or Regulation

Primary Requirement

Key Date

CPPA ADMT regulations

Notice, access, and applicable opt-out rights

January 1, 2027

AB 566

Browser opt-out preference signals

January 1, 2027

AB 1043 / 1856

Age-bracket signaling

Jan. 1 / July 1, 2027

AB 1709

Restrictions on addictive features for minors under 16

Jan. 1, 2027

AB 2246

Children’s privacy and design safeguards

Jan. 1, 2027

SB 1050

Synthetic performer advertising disclosures

Jan. 1, 2027

SB 951

AI-related workforce reduction notices

July 1, 2027

AB 1883

Workplace surveillance restrictions

July 1, 2027

SB 947

Human review of certain employment decisions

July 1, 2027

SB 1119

Companion chatbot child-safety protections

July 1, 2027

AB 1159

Higher education student information protections

July 1, 2027

CPPA risk assessments

Initial summaries and attestations

Apr. 1, 2028

CPPA cybersecurity audits

First submissions for earliest applicable cohort

Apr. 1, 2028

1. New Requirements for Automated Decision-Making Under the CCPA

Beginning January 1, 2027, businesses subject to the CCPA that use automated decision-making technology (ADMT) to make significant decisions concerning consumers must comply with new regulatory requirements adopted by the California Privacy Protection Agency (CPPA). Although the CPPA’s broader regulatory package became effective January 1, 2026, businesses received additional time to implement the ADMT-specific requirements.

Under the regulations, a significant decision generally includes certain decisions concerning employment, education, housing, financial services, and healthcare. Covered businesses must provide consumers with a pre-use notice describing the purposes for which ADMT will be used, applicable opt-out rights, and information concerning their rights to access information about the automated decision. The regulations also establish a right to obtain meaningful information about how ADMT was used to make a significant decision, including information concerning the system’s output and its role in the decision-making process.

Businesses must generally provide an opportunity to opt out of covered ADMT uses unless an applicable exception is satisfied. Certain exceptions depend on the nature of the decision or the availability of a qualifying human review process. For example, a financial services company using an automated system to make lending decisions should determine whether its technology falls within the regulatory definition of ADMT, whether the decision is significant, and whether appropriate notices and consumer rights procedures are in place. These obligations may require businesses to revise privacy notices, implement new request-response procedures, and obtain additional information from technology vendors concerning how automated systems generate and use outputs.1

2. Expanded Privacy Risk Assessment and Cybersecurity Obligations

The CPPA’s regulations also establish requirements for businesses to conduct privacy risk assessments and, where applicable, annual cybersecurity audits. Businesses engaging in certain processing activities that present significant risks to consumers’ privacy must conduct and document risk assessments. Covered activities may include specified uses of sensitive personal information, certain automated processing, and other activities identified in the regulations.

The risk assessment requirements became effective January 1, 2026. Businesses subject to these obligations must submit their initial risk assessment summaries and attestations to the CPPA by April 1, 2028. Separately, businesses meeting the applicable cybersecurity audit thresholds must conduct independent cybersecurity audits addressing their compliance with specified security requirements. The first audit period may begin January 1, 2027, while submission deadlines are phased according to annual gross revenue, beginning April 1, 2028.

Businesses should assess whether existing privacy impact assessments, cybersecurity evaluations, and AI governance procedures satisfy the applicable regulatory standards. A general internal review of AI risks may not be sufficient where the regulations prescribe particular assessment criteria, documentation, or reporting obligations.2

3. Stronger Privacy Protections for Children and Age Verification

California has enacted several laws addressing children’s privacy and online safety, including Assembly Bill (AB) 1043, AB 1856, AB 1709, and AB 2246.

Digital Age Assurance Act (AB 1043 and AB 1856)

Beginning January 1, 2027, covered operating-system providers must establish mechanisms allowing users to provide age-related information and communicate prescribed age-bracket signals to application developers. 

The framework distinguishes users under 13, ages 13 through 15, ages 16 through 17, and adults. Covered application developers must request age-bracket signals under the circumstances specified by the statute. Certain provisions governing existing devices and previously downloaded applications have later transition dates in July 2027. These requirements are intended to facilitate age-appropriate protections without necessarily requiring every application to independently collect a user’s exact date of birth. 

Restrictions on Addictive Features (AB 1709)

Beginning in 2027, covered online platforms are prohibited from providing certain addictive features to users under 16. Restricted features include qualifying personalized feeds and autoplay functionality, subject to statutory definitions and exceptions. Covered platforms must implement reasonable measures to determine whether users are under 16 and prevent access to prohibited features.

Children’s Online Privacy (AB 2246)

AB 2246 establishes additional protections for online services, products, or features likely to be accessed by children under 18. The law addresses privacy-by-default settings, profiling, geolocation, data minimization, and measures intended to reduce specified risks of harm.

Businesses operating consumer-facing websites and applications should evaluate whether children are reasonably likely to access their services, even if the products are not specifically marketed to minors. These laws have distinct coverage thresholds and compliance requirements. Accordingly, businesses should avoid assuming that implementing an age-verification mechanism alone will satisfy all applicable child privacy and online safety obligations.3

4. New Safeguards for AI Companion Chatbots

Senate Bill (SB) 1119, known as Adam’s Law, establishes additional safety and privacy requirements for operators of covered AI companion chatbots. Its principal requirements become operative July 1, 2027. 

The law addresses AI systems designed to engage users in ongoing conversational relationships, particularly where children may interact with the technology. Covered operators that permit children to use companion chatbots must implement specified safety measures, including age assurance, risk assessments, age-appropriate disclosures, parental controls, and safeguards addressing potentially harmful interactions. The statute also establishes requirements concerning suicide and self-harm crisis protocols and restricts certain uses of children’s conversational information. For child users, covered operators generally must implement protective default settings addressing persistent conversational memory, push notifications, and session duration, subject to statutory exceptions and age-specific provisions.

The law further restricts certain targeted advertising practices and the sale or secondary use of information collected from children through companion chatbots. Importantly, the statute does not automatically apply to every AI-powered customer-service tool or workplace chatbot. Businesses must evaluate whether a particular technology falls within the statutory definition of a companion chatbot and whether an exclusion applies. Operators should also evaluate their obligations concerning record retention, incident response, and future independent safety audits. Certain audit requirements are subject to later deadlines and statutory revenue thresholds.4

5. New AI Requirements for Employers and Workplace Surveillance

California has also adopted new legislation addressing the use of automated technologies in employment decisions and workplace monitoring.

SB 947: Automated Employment Decisions

Beginning July 1, 2027, employers are prohibited from relying solely on covered automated decision systems to make disciplinary or termination decisions. Where an employer primarily relies on an automated system in making such a decision, the statute requires independent human corroboration and specified post-use notice to the affected employee. Employees also receive certain rights to request information concerning the data used in the automated decision. Employers should review automated performance evaluation, employee monitoring, and disciplinary systems to ensure that appropriate human oversight is incorporated into decision-making procedures.

This bill is explored in a GRSM article. 

SB 951: Technology-Related Workforce Reductions

Effective January 1, 2027, SB 951 adds disclosure requirements under the California Worker Adjustment and Retraining Notification Act (Cal/WARN) when a qualifying mass layoff, relocation, or termination is caused in whole or substantial part by AI or other technology replacing or automating employee positions. Covered employers must provide additional information concerning the technology displacement, including affected positions and the relevant automation.

AB 1883: Workplace Surveillance

Effective January 1, 2027, AB 1883 restricts employers from using covered workplace surveillance technologies to recognize or infer employees’ emotional states or collect neural data, subject to limited exceptions. These requirements operate alongside California’s existing employment discrimination and privacy laws, including the CPPA’s ADMT regulations. Employers using AI in hiring, performance evaluations, scheduling, or disciplinary decisions should assess the separate requirements that may apply to each use case.

6. Browser-Level Privacy Choices and Consumer Opt-Out Rights

AB 566, the California Opt Me Out Act, becomes operative January 1, 2027, and requires covered businesses developing or maintaining internet browsers to provide consumers with a readily accessible mechanism for configuring an opt-out preference signal. The signal allows consumers to communicate their preference to opt out of the sale or sharing of personal information through browser settings.

The law places obligations on browser developers, while businesses receiving valid opt-out preference signals remain subject to the CCPA’s existing requirements governing recognition and processing of those signals. Businesses involved in digital advertising, analytics, and cross-context behavioral advertising should review whether their websites and consent-management systems appropriately recognize and honor applicable opt-out preference signals. Organizations should also evaluate whether third-party advertising and analytics vendors process consumer preferences consistently with their contractual and statutory obligations.5

7. Restrictions on Student Information and AI Model Training

AB 1159 expands California’s student privacy protections and establishes additional requirements for certain education technology providers. The legislation restricts specified uses of covered student information, including certain uses to train generative AI models or develop AI systems. The law also establishes the Higher Education Student Information Protection Act, which becomes operative July 1, 2027, and imposes requirements on covered technology operators serving higher education institutions.

These requirements address the collection, use, disclosure, and security of student information, including restrictions concerning targeted advertising, profiling, and AI development. Education technology providers should review whether information contained in student accounts, assignments, communications, and platform usage records is used for model training or other secondary purposes. Vendor agreements should also be reviewed to determine whether they adequately restrict the use of covered student information by service providers and subcontractors.6

8. AI-Generated Advertising and Synthetic Content

California has also expanded transparency requirements for certain AI-generated content used in commercial advertising. SB 1050 establishes disclosure obligations for covered advertisements prominently featuring synthetic performers created using generative AI. 

The law generally requires clear and conspicuous disclosures where a covered advertisement prominently features a realistic AI-generated human performance that satisfies the statute’s definition of a synthetic performer. The legislation contains exceptions, including certain advertisements involving expressive works and uses limited to translation or accessibility functions. Businesses using AI-generated spokespersons, voices, or performers in digital advertising should assess whether their marketing materials trigger the disclosure requirements. Marketing and legal teams should also establish procedures for reviewing AI-generated content before publication, particularly where third-party agencies or vendors produce advertisements on the company’s behalf.7

9. Broader AI Transparency, Governance, and Accountability

California’s 2027 developments also build on previously enacted legislation governing AI transparency and accountability. The Transparency in Frontier Artificial Intelligence Act, enacted through SB 53, establishes requirements for certain large AI developers concerning safety frameworks, transparency, and reporting of specified critical safety incidents. 

California has also enacted legislation establishing frameworks for independent AI verification organizations and AI auditors, including SB 813 and AB 1405. These measures contain implementation milestones extending into 2028 and 2029 and should not be treated as imposing identical compliance deadlines on all businesses in 2027. Additionally, California’s AI transparency framework addresses certain disclosures and mechanisms for identifying AI-generated content. Businesses developing or deploying generative AI systems should separately evaluate whether they are subject to these requirements. Together, these developments illustrate California’s increasing emphasis on transparency, accountability, and documented risk management throughout the AI lifecycle.

10. Recommended Compliance Steps for Businesses

Businesses should begin preparing for the 2027 requirements by evaluating their existing privacy and AI governance programs and identifying areas requiring additional controls.

Key considerations include:

  • Inventory AI systems and data flows. Identify automated technologies used in consumer interactions, employment decisions, advertising, education, and other business operations. Determine what personal information each system collects, processes, retains, or shares.

  • Evaluate applicable legal requirements. Determine which California laws and regulations apply to each technology based on the business’s role, the system’s functionality, and the individuals affected.

  • Update privacy notices and consumer rights procedures. Review whether existing disclosures, opt-out mechanisms, and access request procedures satisfy the CPPA’s ADMT requirements and other applicable privacy laws.

  • Review children’s privacy protections. Assess age-assurance mechanisms, privacy defaults, profiling restrictions, chatbot safeguards, and controls governing children’s personal information.

  • Strengthen vendor oversight. Review contracts with AI providers, advertising partners, analytics vendors, and other service providers to address data use, model training, confidentiality, security, and regulatory compliance.

  • Implement appropriate human oversight. Establish procedures for reviewing consequential automated decisions, particularly in employment and other contexts involving significant decisions about individuals.

  • Document compliance efforts. Maintain records of risk assessments, technical safeguards, vendor evaluations, notices, and relevant decision-making procedures.

Because the statutes establish different effective dates, operative dates, and enforcement mechanisms, businesses should prioritize compliance efforts according to the specific requirements applicable to their operations.

Conclusion

California’s expanding AI and privacy framework will introduce substantial new obligations for businesses in 2027. From automated decision-making and consumer privacy rights to children’s online safety, workplace surveillance, and AI-generated advertising, the new requirements reach a broad range of technologies and business practices. The legislation also demonstrates that AI compliance can no longer be addressed solely through general privacy notices or internal technology policies. Businesses must increasingly evaluate how automated systems operate, what information they process, how their outputs affect individuals, and whether appropriate safeguards are implemented.

Organizations should take a proactive approach by assessing their AI technologies, updating relevant privacy and governance practices, and coordinating compliance efforts across legal, information security, human resources, product development, and vendor management teams. Early preparation will help businesses address California’s evolving regulatory requirements while continuing to develop and deploy AI technologies responsibly.

1. Cal. Code Regs., tit. 11, § 7001, 7200; Cal. Code Regs., tit. 11, § 7220-7222.

2. Cal. Code Regs., tit. 11, §§ 7120–7124, 7150, et seq.

3. Cal. Civ. Code §§ 1798.500–1798.504, 1798.99.28–1798.99.34; Cal. Bus. & Prof. Code §§ 22682–22686.

4. Cal. Bus. & Prof. Code §§ 21810–21818. 

5. Cal. Civ. Code § 1798.136.

6. Cal. Bus. & Prof. Code §§ 22584–22587.2; AB 1159.

7. Cal. Bus. & Prof. Code § 17610; SB1050.