Sam Altman shares Anthropic’s concerns when it comes to working with the Pentagon

In a recent discussion, Sam Altman, CEO of OpenAI, articulated the shared apprehensions that both OpenAI and its competitor, Anthropic, have regarding collaboration with the Pentagon. A spokesperson from OpenAI confirmed to CNN that the issues at hand remain consistent across both organizations, particularly concerning the implications of AI in autonomous weaponry and the surveillance of American citizens. During an interview with CNBC, Altman emphasized the necessity for technology firms to engage with the Pentagon, provided that such collaboration adheres to legal standards and respects the critical boundaries outlined by OpenAI and others in the AI field. He acknowledged the differences he has with Anthropic but expressed a level of trust in their commitment to safety and their support for military personnel. Anthropic's Claude AI model has notably been the first to engage with the military’s classified operations. However, the Pentagon has issued a deadline, requiring Anthropic to eliminate its internal safeguards by 5:01 PM on Friday, allowing the AI to be utilized for “all lawful use.” Failure to comply could result in the loss of a significant $200 million contract and a designation as a “supply chain risk,” akin to companies linked to foreign adversaries. Despite its willingness to collaborate, Anthropic has raised alarms about the deployment of AI in autonomous weapons and mass surveillance, citing concerns over the technology's current reliability. Additionally, the company pointed out that existing laws and regulations are ill-equipped to address the rapid advancements in AI technology. In an internal memo reported by the Wall Street Journal, Altman indicated that OpenAI is actively pursuing an agreement with the Pentagon for its models to be utilized in classified settings, which he hopes will also address the unresolved issues between Anthropic and the military.