OpenAI grants US government early GPT-5.5 access for national security
OpenAI is letting the US government try out its latest AI model, GPT-5.5, before public release to help with national security checks.
Announced by Chris Lehane, OpenAI's vice president of global affairs, this move teams up with the Center for AI Standards and Innovation (CAISI) and includes a special Cyber version focused on boosting cyber defenses.
OpenAI says GPT-5.5 outperforms rivals
GPT-5.5 isn't just another upgrade: it's built for tough tasks like independent planning, using tools, and fixing its own mistakes.
OpenAI says it outperformed rivals like Anthropic's Claude Opus 4.7 and Google's Gemini 3.1 Pro in coding and advanced math tests.
Soon, cybersecurity professionals will get access to help protect critical infrastructure, while OpenAI works with the White House to make sure AI is used responsibly at all levels of government.