The White House is considering a significant shift in its approach to regulating advanced artificial intelligence (AI), a move that has sparked debate and concern within the tech industry. This potential policy change comes at a time when public unease over AI is growing, as evidenced by recent poll results.
The AI Vetting Debate
One of the key actions being contemplated is an official pre-deployment vetting process for advanced AI models. This proposal aims to address national security risks and the industry's influence on government policies. The idea is to evaluate AI models for potential threats before they are released to the public.
Executive Order on the Horizon
According to sources, the administration is drafting an executive order that would prohibit private companies from interfering with the government's use of AI models. This order would also introduce stricter contracting and termination standards for federal vendors. The recent standoff between the Defense Department and Anthropic, an AI company, has brought these issues to the forefront. Anthropic's refusal to allow the military to use its AI model, Claude, for surveillance and autonomous weapons has led to a supply chain risk designation, limiting the government's access to their products.
A Policy U-Turn
The proposed tighter controls on AI represent a departure from the White House's previous hands-off approach, which was influenced by laissez-faire venture capitalists. This shift has alarmed tech circles, with industry representatives fearing that increased government intervention will hinder innovation. Daniel Castro, president of the Information Technology and Innovation Foundation, warns that the pace of innovation in Silicon Valley could be slowed down by the need to seek government approval for new AI models.
Cybersecurity Concerns
Another aspect of the executive order focuses on cybersecurity risks posed by advanced AI models, particularly Anthropic's Mythos. Early testing suggests that Mythos can identify and exploit software vulnerabilities in ways that human hackers cannot. The order aims to establish technical guidelines and best practices to secure open-weight models and ensure the safety of AI-powered systems.
Mythos and National Security
The emergence of Mythos has shifted the conversation around AI and national security within the White House. Saif Khan, a former adviser on emerging technology, believes that the potential threat posed by Mythos has led to a more serious approach to AI policy. The administration is now working to create a board to review the supply chain risk designation against Anthropic, indicating a willingness to engage with the industry and address concerns.
Conclusion
The White House's consideration of tighter controls on advanced AI highlights the complex balance between innovation and national security. As the administration navigates this delicate path, the tech industry watches with a mix of concern and anticipation, hoping for a policy that fosters both progress and protection.
