Sam Altman faced 'serious questions' in meeting with lawmakers about OpenAI's defense work

Sam Altman, the CEO of OpenAI, recently engaged in a crucial dialogue with several lawmakers in Washington, D.C., where significant concerns were raised regarding the company’s involvement in defense initiatives. Senator Mark Kelly, a Democrat from Arizona, highlighted that he presented "serious questions" about OpenAI's approach to military applications and its recent partnership with the Department of Defense (DOD). During an interview with CNBC’s Emily Wilkins, Kelly emphasized that the discussions delved into the implications of artificial intelligence (AI) in surveillance and its potential role in military operations. He described the exchange as a "good discussion," asserting the necessity for regulatory measures to ensure compliance with constitutional principles. OpenAI’s agreement with the DOD was finalized last month, shortly after rival firm Anthropic was labeled a "Supply-Chain Risk to National Security" by Defense Secretary Pete Hegseth. This occurred amid stalled negotiations between Anthropic and the DOD over the conditions of their contract, particularly regarding technology usage. The DOD sought unrestricted access to Anthropic’s models for lawful purposes, while the latter pushed for guarantees against the deployment of their technology in fully autonomous weapon systems or for domestic surveillance. In response to the breakdown of Anthropic's discussions, Altman took to social media to reaffirm OpenAI’s commitment to essential safety principles, including prohibitions against domestic surveillance and ensuring human oversight over the application of force. He asserted that these principles were integral to the agreement with the DOD, which allows the agency to utilize OpenAI’s systems for all lawful activities. Altman expressed confidence that the DOD would adhere to these ethical boundaries, citing the contract's language and current laws as protective measures. He underscored the importance of supporting the U.S. government and maintaining a democratic process in technology deployment. The unexpected rift between Anthropic and the DOD has stirred discussions among officials and tech experts, as Anthropic's AI models were previously regarded as superior within classified military networks. Senator Kelly remarked on the necessity for congressional input in establishing legislative frameworks that would provide the necessary boundaries and regulations around AI technology in defense. He acknowledged the challenges of legislative progress, stating, "This is the United States Congress; things don’t move as fast as we would like, but this technology is advancing rapidly."