Archive — US — techcrunch
Anthropic submits AI policy recommendations to the White House
Anthropic submitted AI policy recommendations to the White House after removing Biden-era commitments. The recommendations aim to enhance the economic benefits of AI and include safety measures, national security evaluations, export controls, and power targets for AI data centers by 2027, aligning closely with Biden's previous AI Executive Order repealed by Trump.
Key points
- Anthropic removed Biden-era AI policy commitments from its website.
- They submitted new recommendations for a national AI policy.
- Proposals include maintaining the AI Safety Institute and national security evaluations from NIST.
- Calls for a government team to identify AI security vulnerabilities.
- Advocated for tighter export controls on AI chips, especially to China.
- Recommendation to build 50 gigawatts of power for the AI sector by 2027.
- Suggestions align closely with practices from Biden's AI Executive Order, which was repealed under Trump due to criticisms of reporting burdens.
Why it matters
- Historical context includes President Biden's AI Executive Order, advocating for responsible AI policies.
- Concerns about national security regarding AI technology are heightened as global competition increases.
- The AI industry is expected to shape economic landscapes substantially by 2025.
- Debates around AI regulations are part of broader discussions about tech governance and security.
Anthropic's AI policy recommendations reflect a strategic effort to secure and expand the U.S. leadership in AI technology while addressing national security concerns amidst ongoing political debates over tech regulations.