Question of the Day

The control requirements that separated Anthropic's "security risk" exit from OpenAI's Pentagon re-entry primarily involved the use of AI for surveillance and autonomous weapons. Anthropic was labeled a national-security supply-chain risk partly because it refused to modify its AI technology for mass domestic surveillance and lethal autonomous weapons, which led to the Pentagon cancelling a $200 million contract and labeling them a supply-chain risk [1], [2], [4], [5]. In contrast, OpenAI entered into an agreement with the Pentagon under the terms that included an all-lawful-purposes framework while adopting additional safeguards against the use of AI technology for autonomous weapons and ensuring human oversight, which satisfied the Pentagon's requirements [2], [3], [6]. OpenAI aimed to include these safety guardrails to fulfill government contract stipulations while navigating the restrictions that Anthropic resisted , [5].