Pentagon Issues Ultimatum to Anthropic Over AI Model Claude
DoD demands Anthropic loosen AI safeguards for military use.
Top Summary
- What happened: Anthropic CEO Dario Amodei is summoned to the Pentagon for a critical meeting regarding military use of the Claude AI model.
- Why it matters: Claude is the only AI model currently used in the military's classified systems for sensitive defense and intelligence work.
- What changes for people: Potential changes in AI usage restrictions for national security and defense purposes. Could affect how the US military utilizes AI.
- Who is affected: The US Department of Defence, Anthropic, other AI labs, and potentially other companies working with the Pentagon.
High-Stakes Meeting at the Pentagon
Anthropic CEO Dario Amodei faces a crucial meeting at the Pentagon, summoned by Defence Secretary Pete Hegseth. The meeting is centered on the military's use of Anthropic's Claude AI model.
A senior DoD official described the meeting to Axios as a "sh*t-or-get-off-the-pot meeting."
The Core of the Dispute
The Pentagon seeks greater access to Claude, currently the only AI model within its classified systems. They are reportedly furious with Anthropic's refusal to fully lift its safeguards.
While willing to loosen restrictions, Anthropic wants to maintain safeguards around mass surveillance of Americans and the development of weapons that fire without human involvement.
Anthropic "is committed to using frontier AI in support of US national security," a spokesperson said.
Impasse in Negotiations
DoD officials claim negotiations have stalled, nearing a breaking point. The Pentagon wants AI labs to make models available for "all lawful uses".
The Pentagon has threatened to declare Anthropic a "supply chain risk." This could void contracts and force other companies to certify they aren't using Claude in prohibited ways.
Escalating Tensions
Hegseth will present Amodei with an ultimatum, according to a Defence official. Replacing Anthropic would be difficult due to Claude's deep integration and the current limitations of other AI models.
Officials describe a culture clash between the Pentagon and the Silicon Valley firm.
"The problem with Dario is, with him, it's ideological. We know who we're dealing with," said a senior Pentagon official.
Broader Concerns About AI
Anthropic isn't alone in its concerns. Questions linger about US law's ability to keep pace with AI's surveillance potential and the risks of AI-powered weapons.
The use of Claude in the raid on Venezuela’s ex-president, Nicolas Maduro, further strained relations.
Key Players in the Meeting
Leading the Pentagon side are Hegseth, Deputy Secretary Steve Feinberg, and Under Secretary Emil Michael. Michael has been heading negotiations with Anthropic and three other AI model-makers.
Anthropic has not disclosed its delegation.
What to Watch Next
The outcome of the meeting will determine the future of the US military's access to advanced AI capabilities and potentially influence broader regulations surrounding AI usage in defense. Pay close attention to any announcements or policy changes from both the Pentagon and Anthropic in the coming weeks.