OpenAI Limits Cyber Access After Criticizing Anthropic
SkimNews Take
The industry's move to restrict access to advanced AI cybersecurity tools, even after public criticism of such practices, suggests a tacit agreement on the unique risks these powerful models pose, regardless of competitive posturing.
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI will begin rolling out GPT‑5.5 Cyber to “critical cyber defenders” in the next few days, requiring applicants to submit credentials and intended use.
- Sam Altman previously called Anthropic’s Mythos gatekeeping “fear‑based marketing” before OpenAI adopted a similar restriction.
- Anthropic limited Mythos to a select group of users, prompting an unauthorized group to obtain access despite the restrictions.
- OpenAI is consulting with the U.S. government to identify legitimate cybersecurity users and expand Cyber’s availability beyond the initial rollout.
Why it matters: Critical cyber defenders get early access to Cyber within days, while malicious actors are denied a powerful testing kit; the tighter control steers the AI security market toward government‑approved channels.
