Anthropic, a prominent AI startup, has submitted a comprehensive set of recommendations to the White House for a national AI policy, aiming to "better prepare America to capture the economic benefits" of artificial intelligence. This move comes just a day after the company quietly removed Biden-era AI policy commitments from its website, sparking curiosity about its stance on AI governance.

The company's proposals include preserving the AI Safety Institute established under the Biden Administration, which was a key component of the former President's AI Executive Order. Anthropic also suggests directing the National Institute of Standards and Technology (NIST) to develop national security evaluations for powerful AI models, ensuring that these models are safe and secure for deployment.

Another crucial aspect of Anthropic's recommendations is the establishment of a dedicated team within the government to analyze potential security vulnerabilities in AI systems. This team would focus on identifying and mitigating risks associated with AI adoption, thereby protecting national security interests.

In a move that may raise eyebrows, Anthropic calls for hardened AI chip export controls, particularly restrictions on the sale of Nvidia H20 chips to China. This proposal is likely driven by concerns about the potential misuse of advanced AI capabilities by foreign entities, and the need to safeguard American technological advancements.

In addition to these security-focused recommendations, Anthropic proposes that the U.S. establish a national target of building 50 additional gigawatts of power dedicated to the AI industry by 2027. This ambitious goal would require significant investments in infrastructure and energy production, but could potentially create a competitive advantage for the U.S. in the global AI landscape.

Notably, several of Anthropic's policy suggestions closely align with former President Biden's AI Executive Order, which was repealed in January by the Trump administration. Critics of the order had argued that its reporting requirements were overly burdensome, but Anthropic's proposals seem to suggest that some form of regulation is necessary to ensure responsible AI development and deployment.

The submission of these recommendations marks a significant development in the ongoing conversation about AI governance and policy in the United States. As the AI industry continues to grow and evolve, it remains to be seen how the White House will respond to Anthropic's proposals and whether they will inform future policy decisions.

In the broader context, Anthropic's move highlights the need for a coordinated national approach to AI policy, one that balances innovation with security and accountability. As AI becomes increasingly integral to various aspects of modern life, the stakes for getting policy right have never been higher. The White House's response to Anthropic's recommendations will be closely watched, and could have far-reaching implications for the future of AI in America.