- OpenAI rolled out a model that rivals the capabilities of Mythos — without nearly as much fanfare or as Anthropic received.
- The new GPT-5.5-Cyber achieved an 85.6% score in CyberGym, an internal benchmark that measures whether an AI agent can reproduce known software vulnerabilities.
- In comparison, Mythos 5 scored 83.8% on the same evaluation.
- The White House and OpenAI did not respond to requests for comment.
- OpenAI also said it has expanded its partnerships with organizations in Australia, Canada, France, Germany, Japan, Poland, South Korea and the EU.
OpenAI's latest model, GPT-5.5-Cyber, has emerged as a formidable competitor to Mythos 5, achieving an impressive 85.6% on the CyberGym benchmark, which assesses AI's ability to replicate known software vulnerabilities.1234
In contrast, Mythos 5 scored 83.8%, highlighting OpenAI's advancements in AI technology.
Despite the significance of this development, the White House has remained silent on the matter, with both the administration and OpenAI not responding to requests for comment.
OpenAI has also announced an expansion of its partnerships with various organizations across the globe, including countries such as Australia, Canada, France, Germany, Japan, Poland, South Korea, and the EU. This move indicates OpenAI's commitment to enhancing its collaborative efforts internationally, potentially influencing the future landscape of AI technology.5
As the competition intensifies, the implications of these advancements could reshape the AI industry, prompting further scrutiny from policymakers and stakeholders alike.
“OpenAI has introduced a new model that competes with Mythos, achieving an 85.6% score in CyberGym. The White House has not commented on this development.”
