OpenAI to Allow the EU Access to Newly Launched Cyber AI Model
In Focus
- OpenAI will be rolling out the Cyber AI model in limited preview capacity
- The EU welcomed OpenAI’s intention to allow European entities access to Cyber AI
- Anthropic has not granted EU preview access to Claude Mythos Preview yet
OpenAI is talking to the EU about access to its Cyber AI model. The company said on May 11, 2026 that it will grant European partners access to GPT-5.5-Cyber. OpenAI opens access to its newly launched model to the EU as Anthropic withholds the release of Claude Mythos to the regional bloc.
OpenAI Limits Cyber Access to Preview Capacity
OpenAI will be rolling out the Cyber AI model in a limited preview in the EU. The AI model will be accessible to selected cybersecurity teams in governments, enterprises, cyber authorities, and institutions such as the EU AI office.
“AI labs like ours shouldn’t be the sole arbiters of cyber safety as resilience depends on trusted partners working together. The latest cyber AI capabilities should be available for Europe’s many defenders, not just the few, and we want to help make that happen,” OpenAI’s Head of OpenAI for Countries George Osborne said as cited by CNBC.
The European Commission has confirmed that discussions between the EU and OpenAI about access to the new AI model will be happening in the coming days.
“We welcome OpenAI’s transparency and intent to give the Commission access to the new model. This will allow us to follow deployment of the model very closely, and address security concerns,” European Commission Spokesperson Thomas Regnier noted in a statement.
OpenAI launched Cyber a month after Anthropic’s Claude Mythos sparked fears over hacking risks.
What is OpenAI’s Cyber AI Model?
GPT-5.5-Cyber is a variant of OpenAI’s GPT-5.5, which launched last week. Cyber is specifically designed for high-impact cybersecurity research. GPT-5.5 became ChatGPT’s default model, replacing GPT-5.3 Instant which the company released in March 2026.
OpenAI’s Cyber model is designed to support a range of cybersecurity tasks. It can complete tasks such as malware analysis, vulnerability identification, and patch validation. The AI lab said its partnership with EU entities would support public safety by making its defensive tools widely accessible.
“Through the OpenAI EU Cyber Action Plan, we will work with European policymakers, institutions and businesses by democratizing access to the defensive tools that trusted actors can use to strengthen shared security, support public safety and reflect European priorities.” Osborne added.
How is Access to Anthropic AI Tools in the EU?
Anthropic has not granted the EU preview access to the Mythos cybersecurity model yet. Regnier said the EU is engaging Anthropic about access to Claude Mythos but discussions have not progressed as much as they have with OpenAI.
Reports show that Anthropic was in discussions with the EU about access to AI Code Review and other models that are not accessible in the region last month. Anthropic introduced Claude Mythos Preview in April 2026 as part of a new cybersecurity initiative, Project Glasswing.
The AI model is designed to identify security flaws and vulnerabilities within software. The company is partnering with major U.S. tech firms including Google, Apple, and Microsoft, in the initial phase of the Glasswing Project.
