U.S. Cyber Command to Deploy Top AI Models for Cyber Operations
Similar Articles
AI Firms Brief Congress on Advanced Cybersecurity Models and Risks
White House Developing Guidance to Ease Government Use of Anthropic AI Models
CISA Lacks Access to Anthropic's Security AI Model as Other Agencies Use It for Testing
White House Nears Agreement on AI Safety Oversight Framework
New AI Models Show Advanced Cybersecurity Capabilities in UK Safety Tests
U.S. Cyber Command is building infrastructure to test and deploy the strongest available AI models for cyber operations, regardless of the vendor or country of origin. The command has its first dedicated funding for AI programs this year and is piloting commercial capabilities. Multiple AI companies, including Anthropic, OpenAI, and Google, are now engaging with government agencies to provide their models.
Facts First
- U.S. Cyber Command is building infrastructure to swap between AI models from any vendor or origin.
- 2026 is the first year with dedicated AI funding for Cyber Command, following years of Pentagon and Congressional lead time.
- Multiple AI companies are engaging with government, including Anthropic, OpenAI, and Google.
- Cyber Command is testing varying levels of human oversight but is not considering fully autonomous deployment.
- Access to Anthropic's Mythos Preview model is limited to a patchwork of agencies including the NSA, but not CISA.
What Happened
U.S. Cyber Command intends to test and deploy the strongest possible AI models for cyber operations, according to Brig. Gen. Reid Novotny, the command's first chief AI officer. Cyber Command is using its first dedicated funding for AI programs to pilot commercial AI capabilities and build infrastructure designed to allow operators to switch between models as technology evolves. Multiple AI companies are now engaging with government agencies: [Anthropic] was the first in the classified space, [OpenAI] is engaging with federal, state, and international offices to deploy its GPT-5.4-Cyber product, and [Google] has signed a deal with the Pentagon to use its Gemini model in classified operations.
Why this Matters to You
The integration of powerful AI into national cyber defense and offense could lead to more robust protection for critical infrastructure, which you rely on daily for power, communications, and financial services. This technological push may also accelerate the broader adoption of AI security tools in the private sector, potentially affecting the cybersecurity products available to your employer or for your personal devices. The government's focus on human oversight may help mitigate risks associated with autonomous systems.
What's Next
The White House is negotiating access to Anthropic's Mythos Preview model, which the company has withheld due to its hacking capabilities. Cyber Command will continue to test varying levels of human oversight as it integrates these commercial AI models into its operations. The infrastructure being built to swap between models may allow for rapid adoption of future, more advanced AI systems as they become available.