US Gov to Get Early Look at Advanced AI Models

The White House is poised to issue an executive order that would encourage AI companies to voluntarily submit advanced models for government review prior to public release. This initiative aims to bolster national security and public safety by allowing federal agencies to assess potential threats. Discussions are underway between the government and leading AI firms, including OpenAI and Anthropic, regarding the terms of this proposed review process.

Under the proposed framework, participating AI companies would agree to share their cutting-edge models with the government for a specified pre-launch period. This timeframe is a key point of negotiation, with suggested durations ranging from 14 to 90 days. The objective is to enable the government to identify and address vulnerabilities that could be exploited for malicious purposes, such as sophisticated cyberattacks.

Further details suggest the executive order may include provisions for a cybersecurity "clearinghouse" involving the Treasury Department and other agencies to fix security flaws in unreleased models. Additionally, there's a focus on enhancing the US Tech Force to modernize government computer systems, ensuring readiness for emerging technological challenges.