Americas

Home page

Americas

Pentagon allows use of Grok chatbot in classified systems

24 February 2026 09:26 (UTC +04:00)

The company xAI, founded by Elon Musk, has signed an agreement that allows its Grok model to be used in classified defense systems, APA reports.

Such information was confirmed by a defense official during a discussion with Axios.

According to a source, the agreement opens the way for applying Grok in the most sensitive areas of defense activities – from intelligence and weapons development to planning combat operations. Previously, in similar systems, the Claude model from Anthropic was predominantly used.

However, the Pentagon is pressuring Anthropic in the dispute over safety measures and standards, and may soon require a replacement. This development could alter AI supply chains in the military sphere and influence decisions about the use of specific AI models in government procurement.

The presence of Grok in classified systems signals a new dynamic of interaction between the private sector and government agencies, and underscores the need for clear data-security requirements and ethical use of artificial intelligence in defense.

If the Pentagon reaches an agreement with Anthropic or chooses other solutions, this could spur faster integration of advanced AI technologies into military processes and influence the development of the private AI solutions market for government needs.