OpenAI provided the US government with early access to its GPT-5.5 AI model for national security testing and evaluations, the company disclosed in May 2026. The revelation came via a LinkedIn post by Chris Lehane, OpenAI’s vice president of global affairs.
Lehane confirmed that OpenAI is also working with the US government’s Center for AI Standards and Innovation (CAISI) to test upcoming models, including GPT-5.5 Cyber — a model designed specifically for cyber defenders. OpenAI CEO Sam Altman had separately announced that GPT-5.5 Cyber would be released to “critical cyber defenders” within days, adding that the company wants to “rapidly help secure companies and infrastructure.”
Lehane also outlined a broader partnership with the White House and the wider US administration, describing a “responsible deployment strategy” that includes a playbook for getting AI capabilities into the hands of federal, state, and local governments, allies, and critical infrastructure operators.
The disclosure came on the same day OpenAI announced it was updating ChatGPT’s default model to GPT-5.5 Instant. GPT-5.5, OpenAI’s latest model, is built to handle complex, multi-step tasks with minimal user supervision. It carries what the company calls “agentic” capabilities — the ability to plan tasks, use tools, check its own work, and work through problems independently. OpenAI has claimed the model outperforms Anthropic’s Claude Opus 4.7 and Google’s Gemini 3.1 Pro on several benchmarks, including a score of 82.7 percent on the Terminal-Bench 2.0 benchmark, compared to 69.4 percent and 68.5 percent for its competitors respectively.
The news adds to OpenAI’s growing ties with the US government, which have accelerated since the company launched “OpenAI for Government” last year. The company faced public backlash earlier in 2026 after signing a contract with the US Department of War to deploy AI models in classified work — a partnership announced shortly after reports emerged that Anthropic’s deal with the Pentagon had collapsed over disagreements about safer AI use.
Source: mint – technology