Defense Secretary Pete Hegseth has reportedly set a deadline for Anthropic, requiring the company to remove specific guardrails from its AI model, Claude, by Friday. This ultimatum comes amid mounting pressure from the Pentagon, which is seeking greater military utility from the AI technology. Axios has disclosed that CEO Dario Amodei engaged in discussions with Hegseth yesterday, as the defense establishment intensifies its demands.
Under the terms of the ultimatum, Anthropic faces a critical choice: comply with the government’s request to eliminate restrictions for certain military applications or risk being compelled to modify its AI model under the Defense Production Act. While Anthropic is open to aligning with some Pentagon policies, the company has firmly stated its opposition to using Claude for mass surveillance of American citizens or developing autonomous weaponry.
Key points:
– Deadline imposed by the Pentagon for Anthropic to modify its AI model.
– Discussions between Hegseth and Anthropic’s CEO occurred recently.
– Anthropic willing to cooperate but defiant against mass surveillance and autonomous weapons.
– Claude is currently pivotal in sensitive government projects.
– Pentagon exploring partnerships with OpenAI and Google for classified applications.
The Pentagon is reportedly increasing dialogue with OpenAI and Google regarding the integration of their models into classified operations. It’s also worth noting that ChatGPT and Gemini are already authorized for unclassified government tasks. Furthermore, xAI, founded by Elon Musk, has recently collaborated with the Department of Defense to deploy its Grok system in classified environments.
