Archive — US — techcrunch
Anthropic endorses California’s AI safety bill, SB 53
Anthropic officially endorses California's AI safety bill, SB 53, which requires major AI developers to implement transparency and safety measures before deploying models. Despite pushback from tech groups and differing opinions on federal vs. state regulations, SB 53 aims to mitigate catastrophic risks associated with powerful AI technologies. The bill is awaiting a final vote in the Senate.
Key points
- Anthropic endorses SB 53, a California bill mandating transparency for AI developers.
- SB 53 aims to prevent catastrophic risks, defined as incidents causing at least 50 deaths or over $1 billion in damages.
- The bill requires developers to release public safety reports and provides whistleblower protections.
- Pushback against the bill includes concerns about hindering innovation and interstate commerce.
- Senator Wiener advocates for SB 53 citing safety concerns, while previous AI safety bills faced significant opposition.
Why it matters
- The debate around SB 53 reflects the broader national conversation on AI regulation as states seek proactive governance.
- Historical resistance to technology regulation can be seen in the context of past debates on data privacy and the internet.
- As of 2025, emerging trends in AI safety regulation are shaping discussions about federal versus state-level governance.
The endorsement of SB 53 by Anthropic represents a pivotal moment in AI regulation, aiming to balance innovation with safety, although it continues to spark significant debate among tech stakeholders and policymakers.