OpenAI has launched an updated version of its cybersecurity model, GPT-5.5-Cyber, which scored 85.6% on CyberGym, an internal benchmark for assessing AI's ability to replicate known software vulnerabilities. This release has occurred with less public attention compared to Anthropic's recent model, Fable 5, which faced significant political scrutiny. The White House and OpenAI did not respond to inquiries regarding the differing receptions of the two models. OpenAI has also expanded its partnerships with various organizations across multiple countries, including Australia, Canada, and several European nations. In contrast, Anthropic is currently dealing with export controls that limit the use of its models by foreign nationals, a situation that has been linked to tensions with the Trump administration. The recent G7 Summit highlighted discussions around access to Mythos, Anthropic's model, amid ongoing negotiations and political dynamics.
✓ AI-Debiased Article
4
Wire-neutral
provisional
✓ No loaded language, vague sourcing, or framing detected.
OpenAI Releases New Cybersecurity Model Amidst Political Context
OpenAI has introduced its GPT-5.5-Cyber model, achieving an 85.6% score on CyberGym, while receiving less political backlash than Anthropic's Fable 5. The White House did not comment on the differing reactions to the two models, as OpenAI continues to expand its international partnerships amidst Anthropic's export control challenges.
Companies
OpenAI
Anthropic
People
Donald Trump
Annotating as
No note attached
on this article.
Original vs. Neutral
Original Headline
White House quiet on OpenAI's Mythos-like model
Neutral Headline
OpenAI Releases New Cybersecurity Model Amidst Political Context