California enacts AI law benefiting big tech firms
Governor Gavin Newsom signed a new AI regulation bill into law on September 20, 2025, which critics argue hands significant concessions to major technology companies. The legislation delays stringent safety requirements for AI systems while providing exemptions for large firms. This move comes amid intense lobbying from Silicon Valley.
The bill, known as AB 2013, was introduced earlier in the year to address growing concerns over AI risks, including bias, misinformation, and autonomous decision-making harms. However, after months of negotiations, the final version signed by Governor Newsom includes provisions that tech giants like Google, Meta, and OpenAI had advocated for. According to the Ars Technica report, the law postpones mandatory impact assessments for high-risk AI applications until 2028, giving companies more time to comply or innovate around regulations.
Key elements of the law include exemptions for AI models with over 100 million parameters from immediate transparency reporting, a threshold that covers most products from big tech. 'This is a win for innovation,' said a spokesperson for the California Chamber of Commerce, which supported the bill. The legislation also establishes a state AI advisory council but limits its enforcement powers, focusing instead on voluntary guidelines.
Background context reveals a contentious legislative process. Initial drafts in the California Assembly called for stricter rules, such as real-time auditing of AI outputs and liability for developers in cases of harm. Tech industry groups, including the Information Technology Industry Council, lobbied heavily, arguing that overly burdensome rules could stifle competition and drive jobs out of state. By August 2025, amendments softened these measures, leading to bipartisan passage in the legislature.
Critics, including consumer advocacy groups like the Electronic Frontier Foundation, expressed disappointment. 'This law essentially gives big tech a free pass while smaller innovators face the full brunt,' stated EFF policy director Jeremy Gillula in a statement. The report notes that the exemptions could exacerbate existing power imbalances in the AI sector, where dominant players already control vast data resources.
Implications for the future are significant. The law sets a precedent for U.S. AI governance, potentially influencing federal efforts. As California often leads on tech policy, this diluted approach might encourage other states to follow suit, prioritizing economic growth over rapid safety implementations. No immediate legal challenges have been announced, but watchdogs plan to monitor compliance closely.