The Point: Anthropic and OpenAI pursue opposing strategies for AI regulation at the state level: Anthropic supports progressively stricter standards, while OpenAI seeks uniform regulatory frameworks.
Anthropic is pursuing a different regulatory model than its competitor OpenAI: Instead of pushing for uniform state regulations, the AI safety-focused company supports the concept that states should establish increasingly stringent standards. This has significant implications for future U.S. AI regulation.
Anthropic is pursuing a deliberately different lobbying strategy in U.S. states than OpenAI. While OpenAI representatives promote so-called “reverse federalism” – a national AI regulatory approach through harmonization of individual state laws – Anthropic explicitly rejects this approach. Cesar Fernandez, Head of U.S. State and Local Government Relations at Anthropic, told Politico that the company does not want to “support the same legislative proposal in every single state,” but rather deliberately seeks regulations that significantly increase safety requirements for the most powerful AI systems.
This contrast is strategically relevant for compliance officers: While OpenAI lobbyist Liz Bourgeois argues that uniform standards enable the allocation of resources to security rather than compliance with competing requirements, Anthropic sees the necessity to keep pace with rapid AI model development. Anthropic was the only leading AI lab that supported California’s 2025 law regulating advanced AI models – the first law of its kind in the United States. Additionally, the company supported more ambitious legislative proposals in New York, Illinois, and Massachusetts.
Fernandez justified this position by stating that the rapid development of increasingly capable models requires stronger measures. Anthropic has found in tests of its Claude model that it could potentially exploit security vulnerabilities in common operating systems. In Anthropic’s view, compliance requirements are no longer limited to transparency and self-reporting obligations, but must include substantial security commitments from providers. For organizations in the regulatory environment, this means that fragmentation of AI security requirements across states should be expected.
Source: www.politico.com · Published 15 July 2026
Lumi AI News — AI-assisted curation according to Art. 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.7.3.