Transparency Coalition reported on 4 September that California lawmakers concluded their 2026 session near midnight on Monday having passed 30 AI-related bills, with Governor Newsom holding until 30 September to sign or veto. The measures span chatbot safety, education, medical authorisation, consumer rights and frontier model oversight, and the organisation's mid-year count records 85 new AI-related laws enacted across 27 states so far in 2026, with Michigan, Pennsylvania, Massachusetts, Ohio, New Jersey and North Carolina still sitting.
KQED reported that the bills reaching the Governor include a four-year moratorium on the manufacture and sale of AI chatbot powered toys, alongside measures on data centres and surveillance. OpenAI has publicly urged Newsom to sign the chatbot safety measure, arguing it establishes meaningful safeguards while preserving young people's ability to use AI for learning and creativity.
Regulation
The California Legislature passed Senate Bill 1119, to be known as Adam's Law, on 31 August 2026, with Assembly amendments concurred in by 39 votes to nil and the bill ordered to engrossing and enrolling before transmission to the Governor in the first week of September. The measure adds a new chapter to the Business and Professions Code imposing a duty on operators to ensure a companion chatbot does not pose an unreasonable risk of a covered harm to a minor, requiring annual child safety risk assessments that weigh the likelihood of harm, a public-facing child safety policy describing mitigation measures, and independent third-party audit of the internal controls, processes and systems relied on for compliance, with the operative obligations beginning on 1 July 2027 and repeating before deployment of any new or substantially modified system.
Cases
Via the notice of the Bundesgerichtshof in case I ZR 281/25, it can be seen that the First Civil Senate heard the photographer's appeal on 3 September 2026 and has set judgment for 17 December 2026. The dispute concerns a non-profit association that downloaded a stock agency preview of the claimant's photograph in the second half of 2021 while assembling a public dataset of 5.85 billion image-text pairs usable for training generative AI. The Hamburg Regional Court and the Hanseatic Higher Regional Court both dismissed the claim, holding that although the download reproduced the work under section 16(1) UrhG, it was justified by the text and data mining exception in section 44b UrhG and the scientific research exception in section 60d UrhG. The appellate court reasoned that checking whether an image matched its stored description was itself the extraction of information, so it did not matter whether later model training also qualified as mining, and that the agency's English-language prohibition on automated programs and bots was ineffective because the claimant had not shown it was machine-readable at the relevant time.
Academia
arXiv hosts 'GPS-Bench: A Governance Policy Benchmark for Automating Policy Analysis' by Linh Le, Melanie Bui, My Chiffon Nguyen, Zachary Schlosser and David Williams-King, posted in early September 2026. The benchmark links policies to affected actors and downstream impacts using legislative records, lobbying disclosures, regulatory documents and corporate filings, reconstructing each actor from the dated record so that a simulated position carries provenance rather than being prompted as an archetype, with the 2022 advanced chip export controls used as a worked example.
arXiv hosts 'An Instrument to Evaluate Governance Proposals: AI Policy Analysis at Scale', setting out a jurisdiction-agnostic framework for assessing AI governance proposals through empirically grounded rubrics. The authors argue that policy debate collapses into binary positions that hide the underlying trade-offs, and benchmark commercial language models against a domain-trained rubric-calibrated model for structured analysis of this kind.
Events
TAB Global will hold the World AI Regulation Summit 2026 at The Pelligon in London on 5 and 6 November 2026, with sessions on the operationalisation of South Korea's AI Basic Act and speakers including Gabriele Mazzini, lead author of the EU AI Act.
The National University of Singapore Faculty of Law will host the eleventh International Legal Ethics Conference from 9 to 11 December 2026 on the theme of continuity and transformation, examining how integrity, independence and accountability withstand disruption from legal technology and AI.
Sources: Transparency Coalition, KQED, California Legislature, Bundesgerichtshof, arXiv, TAB Global, National University of Singapore