Anthropic releases ‘Claude Gov’ AI models for US national security customers

The OpenAI rival artificial intelligence (AI) startup Anthropic has introduced a new line of AI models specifically developed for US national security agencies.

This initiative marks a strategic move by Anthropic to strengthen its presence in the public sector and diversify its revenue streams.

Dubbed "Claude Gov," these AI models are custom-built versions of Anthropic’s Claude AI, designed to meet the unique and high-stakes demands of military and intelligence operations.

According to the AI company, the Claude Gov models were shaped using direct feedback from government clients to better support mission-critical applications such as strategic planning, operational assistance, and intelligence analysis.

In a blog post, Anthropic stated that these models are already being utilised by top-tier national security agencies and are accessible only within classified environments.

“These models underwent the same rigorous safety evaluations as all our Claude models,” the company stated.

“They’re purpose-built to handle sensitive information and deliver meaningful insights in secure settings.”

The company also noted that the Claude Gov models are designed to engage more effectively with classified material, showing reduced refusal rates when prompted with sensitive content.

The models also feature enhanced language processing capabilities across dialects vital to national security and have improved analytical skills for interpreting complex cybersecurity data.