The Future of AI Regulation and Legislation: 5 Key Takeaways


1.5 hour watch | September.23.2024

We were honored to host a conversation about the future of AI policy last week in Orrick’s San Francisco office -- and would like to share five takeaways with our clients and friends. The conversation featured:

You can watch the full conversation below.

Key Takeaways

  1. Global AI legislation is emerging quickly and varies widely. Regions like the EU, U.S. and China are adopting different regulatory frameworks and approaches to AI regulation (14:58). Common threads include emphasizing transparency, accountability, safety, ethics and protecting human rights, said Yavorsky, co-leader of Orrick’s AI practice (17:47).
  2. Several U.S. states may regulate AI – and companies can shape those proposals. A dozen states will seriously consider or pass comprehensive AI legislation in 2025, predicted Kudon, a leading strategist for high-stakes legislative campaigns (1:20:22). He strongly recommended that AI companies engage to help shape the bills rather than wait for them to pop up on a legislative tracker (1:21:01).
  3. California legislators are considering an array of bills related to AI. SB 1047 is not the only AI bill on Gov. Gavin Newsom’s desk. SB 942 (the California AI Transparency Act) would transform consumer disclosures and awareness in the fight against deepfakes, Kemp said (27:11).
  4. AI may bring big changes to government and business. Randolph stressed the critical role Bay Area companies play in funding and growing AI-focused businesses (55:03). Randolph also said California’s government is laying the groundwork to better deliver public services and equip the workforce of the future.
  5. The future includes a mix of threats and opportunities. Thomas identified the biggest AI threats and economic opportunities AI represents, especially in tech-focused regions like Silicon Valley. Thomas also outlined his legislative priorities to ensure California businesses remain competitive during a time of rapid growth in AI development and deployment (59:01).