Anthropic’s CEO has strongly criticized rival company OpenAI for its recent decision to secure a defense contract with the United States government, alleging that OpenAI misrepresented the circumstances of the deal. In a communication to employees as reported by The Information, Dario Amodei, Anthropic’s co-founder and CEO, labeled OpenAI’s approach to its agreement with the US Department of Defense as “safety theater.” He emphasized that while OpenAI focused on appeasing its staff, Anthropic prioritized preventing potential misuse of its technology.
The conflict arose after Anthropic’s negotiations with the US Department of Defense for an extended contract, aiming to broaden the military’s access to the company’s AI systems, failed. Anthropic, a previous recipient of a $200 million defense contract, sought explicit guarantees that its AI models would not be utilized for domestic mass surveillance or fully autonomous weapons. However, the Department opted to proceed with OpenAI instead.
OpenAI, in a recent blog post, asserted that its systems could be applied for “all lawful purposes” and clarified that the Department had confirmed the illegality and absence of plans for mass domestic surveillance. The company highlighted that this understanding was clearly outlined in its contract.
In his internal communication, Amodei denounced OpenAI’s public declarations as “blatant falsehoods,” accusing Altman of portraying himself as a mediator while downplaying substantial differences in the companies’ stances. Anthropic’s apprehensions were reportedly centered on the Pentagon’s requirement for its AI tools to be available for any lawful use, a condition that could be subject to change with evolving legal standards.
Following the disclosure of OpenAI’s defense contract, there was a notable decline in downloads of its ChatGPT application, while Anthropic experienced an improvement in its app rankings.