Mumbai, Feb 24 (NationPress) — US Secretary of Defense Pete Hegseth has reportedly called upon the CEO of AI company Anthropic, Dario Amodei, to the Pentagon for a "high-stakes", "tense" discussion regarding the military's utilization of the firm's Claude AI model. This follows Elon Musk's criticism of Anthropic for allegedly appropriating training data.

An Axios report, citing an anonymous senior defense official, stated that the gathering was "not cordial" since Anthropic has been resistant to lifting restrictions on their technology, despite Hegseth's urging for AI companies to do so.

Currently, Claude is the sole AI system operational within classified defense networks, stemming from a $200 million pilot contract established last year. However, in a memo dated January 9, Hegseth requested AI companies to renegotiate terms to eliminate restrictions on their technologies.

Despite this, Anthropic has firmly maintained its position against fully removing safeguards, which include limitations on mass surveillance of American citizens and the creation of fully autonomous weaponry, according to the report.

Defense officials have cautioned that Anthropic may face designation as a "supply chain risk," which could invalidate contracts and limit other Pentagon partners from utilizing Claude, even though a spokesperson for the AI firm described the discussions as "productive." The challenge of replacing Anthropic is significant due to its extensive integration into defense systems, the report further noted.

Simultaneously, the Pentagon is finalizing agreements with Elon Musk’s xAI and is close to a deal with Google for the development of the Gemini model, adding pressure on Anthropic, as reported by The New York Times.

Prior to the meeting, Anthropic claimed that three Chinese AI companies had exploited chatbots to extract millions of Claude outputs to enhance their own models.

In response, Musk criticized Anthropic on X, asserting, "Anthropic is culpable of massive-scale data theft and has incurred multi-billion-dollar settlements for their infractions." He further labeled the AI company as ‘MisAnthropic.’

Claude, developed by Anthropic, is a cutting-edge AI assistant designed for safety, accuracy, and security. This innovative assistant has the potential to automate tasks such as legal document reviews, compliance checks, sales planning, marketing campaign assessments, financial reconciliations, data visualizations, SQL-based reporting, and enterprise-wide document searches.

aar/na