OpenAI Restricts GPT-5.5 Cyber Access To Select Security Professionals

OpenAI has announced a phased rollout for its latest cybersecurity testing tool, GPT-5.5 Cyber, but the launch comes with significant strings attached. Despite previously criticizing competitors for limiting access to advanced models, the company has decided to restrict the initial release of this specialized tool exclusively to a group it defines as "critical cyber defenders."

The decision marks a pivot toward a more guarded distribution strategy for high-stakes AI tools. By prioritizing established security professionals, OpenAI aims to prevent the model's potent capabilities from being weaponized by bad actors. However, the move has already drawn comparisons to industry rivals who have faced similar scrutiny for creating "walled gardens" around their most powerful technologies.

As the cybersecurity landscape becomes increasingly reliant on automated defense, the industry is watching to see how OpenAI defines its list of approved users. Whether this restrictive approach will eventually thaw to include broader research communities remains to be seen, as does the impact of such gatekeeping on open-market innovation.

This story was first reported by TechCrunch.