OpenAI grants European companies access to latest AI models through cybersecurity program

OpenAI is handing the keys to its most advanced cybersecurity AI model to a select group of European companies. The goal: help telecoms, banks, and public services defend themselves against increasingly sophisticated digital threats.

The program, called “Trusted Access for Cyber,” rolls out access to GPT-5.5-Cyber, a specialized model designed specifically for defensive security operations like code analysis and threat-hunting.

Who’s getting access and why it matters

Deutsche Telekom, BBVA, Telefónica, Sophos, and Scalable Capital are among the companies selected for the program, which was reported on May 13, 2026.

The European Commission has also been provided access to OpenAI’s cybersecurity features. OpenAI’s competitors, particularly Anthropic, have been slower to build out equivalent enterprise cybersecurity offerings in Europe.

How GPT-5.5-Cyber actually works

OpenAI says the program is designed with safeguards that limit the model to defensive work. GPT-5.5-Cyber can help a security team analyze suspicious code, hunt for threats across networks, and identify potential attack vectors, but it is not designed to help anyone write exploit code or build offensive tools.

What this means for digital asset markets

Several participants in the Trusted Access for Cyber program operate platforms that touch digital assets directly. BBVA has been active in regulated digital asset services across Europe. Scalable Capital operates a financial platform that handles investment products in a regulatory framework that increasingly intersects with tokenized and digital assets.

The concern about AI models being used to create exploit code poses real dangers to financial institutions and digital-asset custodians alike. Advanced AI models capable of analyzing smart contracts and blockchain infrastructure could, if poorly constrained, help attackers identify exploitable vulnerabilities in DeFi protocols or digital-asset custody systems.

If advanced AI security tools remain gated behind selective enterprise programs, the security gap between large incumbents and smaller crypto firms could widen considerably.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.