Search
OpenAI Releases Cybersecurity Model to Limited Group, Competing with Mythos
1 week ago /
Read about 0 minute
Author：小编
On April 14, 2026, OpenAI announced the launch of the GPT-5.4-Cyber model for specific users. This model focuses on discovering software vulnerabilities and aims to help organizations address issues. GPT-5.4-Cyber will be provided to some participants in OpenAI's 'Cybersecurity Trusted Access' program, initially tested by hundreds of users and expanding to thousands in the coming weeks. The week before, OpenAI's competitor Anthropic released a limited version of the AI tool Mythos, which also focuses on identifying and exploiting software vulnerabilities.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
