Dario Amodei, the CEO of anthropic, is once again negotiating with the Pentagon, trying to save a key military contract on artificial intelligence in the light of extreme criticism and calls of national security. Its reversal comes after a controversial break with last Friday, when President Donald Hyatt ordered federal agencies to stop using the Claude models of Anthropic, and Defense Secretary Pete Hegseth warned that the company posed a threat to the supply chain.
Deal Breakdown
The talks went badly due to the provisions held by Anthropic that discouraged the deployment of Claude to spy on domestic matters through examination of bulk data bought or the functioning of wholly autonomous weaponry. One of the staff memos that was obtained by the Financial Times states that Amodei has romanticized the latest proposal of the Pentagon as a trap by viewing it as reflecting their deepest fears.
Amodei was subsequently denounced as a liar with a God complex on X by Under-Secretary Emil Michael, and Hegseth set a Friday, 5.p.m deadline in which the person would have to be completely unfettered.
The $200 Million contract
The $200-million-dollar deal, which was the first critical capability of Claude to enter into the classified networks, helped the U.S. fight in the Iran war just before the ban. At the same time, OpenAI was awarded a similar contract, which raised criticism – the number of Claude applications downloaded climbed, and ChatGPT fell by uninstall rates by 7.
Industry Heat
A technology association involving Nvidia and Google approached Hegseth to change his stance, stating that labeling an American-based company that was potentially dangerous could have far-reaching consequences.
Anthropic originated in 2021, when the former OpenAI staff kept a philosophy of safety-first, regardless of consistent questioning by government officials. The stance taken by the company was dubbed a mistake by FCC Chair Brendan Carr, and he reiterated that the Department of Defense regulation governs all applicable technologies.
What’s Next
Amodei will have a last meeting with Michael to discuss a compromise with related conditions, as those supported by the CEO of OpenAI, Sam. However, the success of this is limited due to the fact that the risk designation is still being disputed by Anthropic and the fact that the disruption of operations was also a concern of the experts.
A major conflict over this period would shift over $200 million dollars of AI spending on the defense market to other vendors, benefiting OpenAI and at the same time undermining the use of ethical AI in the military. The Department of Defense has the potential to tap and utilise the Defense Production Act by the end of the week, and therefore, close monitoring is justified.