Quick overview

---

Dario Amodei, the head of Anthropic PBC, has resumed talks with the Pentagon regarding the US military’s use of its AI models, potentially resolving a conflict that has captured Silicon Valley’s attention.

 

Amodei had been negotiating a contract governing the Pentagon’s access to Anthropic’s technology with Emil Michael, the undersecretary of defense for research and engineering. However, negotiations broke down last week after the startup demanded guarantees that its AI would not be used for autonomous weapon deployment or widespread surveillance of Americans.

Anthropic was then designated as a supply-chain risk by Defense Secretary Pete Hegseth, a label usually applied to US adversaries.

The military could use Anthropic’s AI again if both parties reach a new agreement, which would also reduce the possibility that the Pentagon would formally blacklist the company. Additionally, it might make rival ef more difficult.

Anthropic, which is currently valued at $380 billion, is expected to generate nearly $20 billion in revenue annually, more than doubling its run rate from late last year. However, the company’s prospects have been clouded by the Pentagon dispute. It is unclear how the Pentagon’s announcement will affect Anthropic’s long-standing primary business of selling to corporate clients in the long run. Meanwhile, it’s becoming more popular among regular users.

Recently, Anthropic’s primary app surpassed Apple download charts, indicating a spike in the company’s popularity. Amodei also garnered support from a large portion of Silicon Valley. Tech organizations that represent big businesses like Google and Apple are pleading with President Donald Trump to reevaluate the designation of Anthropic as a national security threat, claiming that doing so would have negative repercussions for the industry as a whole.