Introduction
California is leading the charge in regulating artificial intelligence (AI) through comprehensive legislation. With 19 bills enacted after reviewing nearly 50 proposals, the state addresses critical aspects of AI, such as privacy, transparency [2] [3] [4], ethical use [4], and healthcare applications. This proactive approach aims to manage AI’s implications across various sectors effectively.
Description
California is at the forefront of regulating the use and deployment of artificial intelligence (AI) tools [1], having enacted a total of 19 bills related to AI after reviewing nearly 50 proposals. These legislative measures address various aspects of AI, including privacy, transparency [2] [3] [4], ethical use [4], and healthcare applications, reflecting a proactive approach to managing the implications of this technology across multiple sectors [4].
A key piece of legislation, AB 2885 [4], establishes a uniform definition of AI, providing clarity for future regulations and ensuring consistency across industries [4], effective January 1, 2025 [2] [4]. AB 2013 mandates that developers of generative AI models disclose the datasets used for training [4], enhancing transparency and addressing concerns about bias [4], with implementation set for 2026 [4]. Additionally, SB 942 requires AI-generated content to be marked with watermarks [4], helping consumers differentiate between human and AI-created materials [4], and calls for the development of public detection tools. This aligns with the newly enacted AI Transparency Act, which also establishes disclosure and notice requirements for AI system providers and sets licensing obligations for third parties utilizing AI systems [3].
In the realm of public trust, SB 896 introduces oversight for AI use in state agencies [4], requiring risk assessments and disclosures regarding AI in public interactions [4]. This is complemented by AB 1008, which amends the California Consumer Privacy Act to enhance privacy obligations for AI systems handling personal data [1], reinforcing consumer protections against the misuse of sensitive information [4]. The legislation also addresses key privacy challenges associated with generative AI, emphasizing purpose limitation [3], data proportionality [3], and business continuity [3].
Legislation specifically targeting AI-generated pornography includes AB-1831 [4], which expands child pornography laws to cover AI-generated content [4], and SB-926 [4], which criminalizes AI-generated nude images used for blackmail [4]. Furthermore, SB-981 mandates social media platforms to implement reporting systems for deepfake content [4], while social media companies are now required to submit semiannual reports on their use of AI in content management.
In the healthcare sector, the Physicians Make Decisions Act (SB 1120) [2], effective January 1 [2] [4], mandates that AI-generated algorithms used by insurers for prior authorization and coverage decisions must consider a patient’s medical history and clinical situation, ensuring that human physicians retain ultimate decision-making authority [2]. This law addresses concerns about potential biases in AI algorithms [2], which can lead to inequitable treatment outcomes [2], particularly for marginalized populations [2]. Additionally, AB 3030 requires that any patient communications generated by AI disclose their AI origin unless reviewed by a licensed healthcare provider [2], enhancing transparency in patient interactions [2].
In the entertainment sector [4], AB-2602 requires studios to obtain consent from actors for AI-generated replicas [4], and AB-1836 extends this requirement to deceased performers [4]. AB 3030 also mandates healthcare providers to disclose the use of AI tools in patient care [4], ensuring informed consent and fair use of AI in patient management.
To promote AI literacy [4], AB 2876 establishes standards for responsible AI use among K-12 students. Lastly [4], AB 2905 requires telemarketers to disclose the use of AI-generated voices in robocalls [4], enhancing transparency in commercial communications [4].
Conclusion
California’s comprehensive approach to AI legislation not only emphasizes consumer rights and ethical standards but also sets a significant precedent in the governance of emerging technologies. As states increasingly regulate AI in healthcare [2], the potential for future legislative actions is anticipated [2], particularly in light of ongoing concerns about the impact of AI on patient care and insurance practices [2]. A webinar featuring experts will address best practices and compliance strategies related to the evolving AI landscape in California [3], focusing on effective data governance, version control [3], and comprehensive model documentation to ensure compliance and mitigate associated risks [3]. Other states like Colorado and Utah are also enacting their own AI laws, reflecting a growing trend in AI regulation.
References
[1] https://www.jdsupra.com/legalnews/a-view-from-california-one-important-8750570/
[2] https://www.medscape.com/viewarticle/new-state-law-will-restrict-ai-prior-authorization-coverage-2024a1000krq
[3] https://www.onetrust.com/resources/californias-approach-to-ai-unpacking-new-legislation-webinar/
[4] https://www.jdsupra.com/legalnews/what-can-we-learn-from-california-s-new-3806826/
