AI Trends for 2026 – Return of the Brussels Effect: AI Transparency Requirements Come to California

Recently-enacted AI transparency regulations took effect in California on January 1, 2026. To enterprises with a global footprint, these California regulations will bear a striking resemblance to transparency obligations under the EU AI Act. Although these laws differ from EU law, examining them together enables enterprises to implement globally-applicable AI compliance programs. Notable California AI transparency obligations include the following:

• SB 243, Companion Chatbots requires that deployers of companion chatbots provide a clear and conspicuous notification that the companion chatbot is not human if a reasonable person would be misled. SB 243’s obligations resemble Article 50 of the EU AI Act, which requires that providers inform individuals that they are interacting with an AI system.
• AB 2013, Generative Artificial Intelligence: Training Data Transparency requires developers of generative AI systems to publish documentation describing training data, including sources of datasets, types of data points, and use of synthetic data in development. Article 53 of the EU AI Act imposes a similar requirement, obligating general-purpose AI model providers to publish a detailed summary of the content used for training the model.
• SB 942, California AI Transparency Act requires covered providers to make available an AI detection tool enabling users to assess whether content was created by the provider’s generative AI system. Covered providers must also include a latent disclosure in AI-generated content and provide users with the option of including a manifest disclosure in their content. Article 50 of the EU AI Act requires that providers of AI systems generating synthetic content ensure that outputs are marked as artificially generated. Originally scheduled to take effect on January 1, 2026, AB 853 delayed the effective date of the California AI Transparency Act until August 2, 2026, bringing it into alignment with the corresponding requirements under the EU AI Act. However, the EU has more recently proposed delaying these requirements until 2027—it remains to be seen whether California will do the same.
• SB 53, Transparency in Frontier Artificial Intelligence Act requires that frontier developers—regardless of size—publish transparency reports when deploying a new or modified frontier model. Large frontier developers must also publish a “frontier AI framework” describing how the developer incorporates industry-consensus best practices and mitigates the potential for catastrophic risks. These requirements align with the risk-based obligations and consideration of industry standards set forth in EU AI Act Articles 6, 56, 67, and 95.

While these California disclosure requirements impose obligations on enterprises, they also enable AI developers and deployers to benchmark against others in the industry. AI developers already exchange safety-related best practices through forums like the Frontier Model Forum. But for AI deployers, these transparency requirements may provide new insights into the tools and safety practices employed by developers and other deployers. As these compliance practices become publicly available, AI developers and deployers should monitor evolving trends in their industries.