Will States Lead the Way on AI Regulation?

Last year was​ a busy time for lawmakers and lobbyists concerned about AI, particularly​ іn California, where Governor Gavin Newsom signed​ 18 new​ AI laws, while also vetoing high-profile legislation. 2025 could see even more action, particularly​ at the state level, according​ tо Mark Weatherford, who has extensive experience​ іn cybersecurity and policy-making.

The Challenge of Keeping Up with Rapid AI Advancements

Weatherford​ іs optimistic that lawmakers will eventually gain the necessary understanding​ оf​ AI​ tо make informed decisions. However, the pace​ at which​ AI evolves presents​ a challenge.​ AI issues discussed just​ a month ago can already​ be outdated.​ He also noted that​ a task force​ іn the​ US House​ оf Representatives released​ a 230-page report after​ a year​ оf work, illustrating how slow the process can be.

State-Level AI Legislation: A Growing Trend

Weatherford believes more regulatory action will occur at the state level than federally. California, for example, passed 12 AI-related bills in recent months. With over 400 AI-related bills introduced across US states in the past year, significant state-level legislative activity is expected. One major concern is the need for “harmonization”—ensuring regulations align across states to avoid confusion and unnecessary complexity for businesses.

California’s Influence on AI Legislation

Weatherford acknowledges that California often leads the charge in tech legislation, setting a precedent for other states. The state has taken a comprehensive approach to AI regulation, balancing privacy concerns with innovation requirements. While Governor Newsom vetoed a major bill, it’s likely that California will introduce even stricter regulations in 2025.

Federal Regulation: Progress, but Slow Pace

While the federal government has shown interest in AI regulation, Weatherford does not foresee major federal legislation in 2025. The focus on less regulation and the differing priorities of various lawmakers might delay progress. However, Weatherford believes that bipartisan support for issues like privacy and cybersecurity could help push forward some form of AI regulation.

The Future of Synthetic Data in AI Regulation

Weatherford is also confident that synthetic data will play a significant role in AI’s future. As the pool of available data shrinks, synthetic data can ensure privacy, eliminate bias, and address non-technical issues like fairness. He believes this is key to enabling more ethical and secure AI development, particularly as regulation becomes more stringent.

Balancing AI Censorship and Regulation

AI censorship, a concern voiced by figures in Silicon Valley, is likely to be regulated. Weatherford emphasizes the importance of finding a balance between reasonable content moderation and restrictive censorship, a task that the government will likely take up in response to societal concerns. The incoming administration’s preference for “less regulation” complicates this, but guidance on AI censorship should come through executive orders or interagency coordination.

The Role of Legislation in Shaping AI’s Future

Despite concerns about AI’s potential risks, Weatherford remains optimistic about the technology’s future. He stresses that clear communication and regulation are necessary to help the public understand AI’s implications while also ensuring responsible development. Although there may be bumpy years ahead, legislation will be crucial in both fostering innovation and protecting society from AI’s potential dangers.