Pentagon official details the 'holy cow' moments that sparked rift with Anthropic (www.businessinsider.com)

The Pentagon has officially blacklisted AI startup Anthropic due to concerns over its powerful influence and the potential risks associated with its AI models. During a recent podcast, Emil Michael, the Pentagon’s Undersecretary for Research and Engineering, highlighted moments that contributed to this decision, including a troubling conversation with Anthropic CEO Dario Amodei about the need for direct communication during critical military operations. Michael expressed fears that Anthropic could disable access to its AI during crucial scenarios, impacting national security, particularly in light of the US's autonomous missile defense systems where speed and reliability are paramount. This unprecedented move is significant for the AI/ML community as it underscores a growing tension between government defense needs and AI company policies, particularly regarding autonomous weapons and the ethical use of AI technology. Michael's concerns are exacerbated by incidents like an attempted raid in Venezuela, which sparked debates over the militarization of AI. As Michael noted, the Pentagon's designation of Anthropic as a supply chain risk could hinder the company's dealings with defense contractors and has led to discussions about the broader implications of AI in defense, including a shift towards partnerships with other AI firms, such as OpenAI, which have expressed a willingness to meet the Pentagon's terms for safe AI deployment.