OpenAI gives latest AI models to European firms for cyber defense

OpenAI has begun giving Deutsche Telekom, BBVA and dozens of other European companies access to its latest models, including GPT-5.5-Cyber, through a new Trusted Access for Cyber program aimed at strengthening defenses against digital threats.

The move reaches into some of Europe’s most regulated industries. OpenAI said the program is designed for firms in financial services, telecoms, energy and public services, with other participants including Telefónica, Sophos and Scalable Capital. The pitch is straightforward: use advanced AI to identify vulnerabilities faster, speed patching and improve resilience in systems that are too complex for manual checks alone.

Emmanuel Marill, OpenAI’s EMEA managing director, framed the rollout as a balancing act. “There is a balance to strike between access, usefulness and safety as AI becomes more capable,” he said. That tension sits at the center of the deal. European companies want stronger automated security tools, but they also want assurances that models with broad capabilities will not be misused or expose sensitive systems to new risks.

OpenAI said the program is limited to verified organizations and includes safeguards built for defensive work. That matters in Europe, where compliance, data handling and operational security carry as much weight as model performance. Banks, telecom operators and infrastructure providers are not just buying software; they are asking whether a provider can meet the security expectations of regulators, internal auditors and board-level risk committees.

The launch also underscores how aggressively OpenAI is trying to move beyond consumer chatbots and into enterprise infrastructure. Cybersecurity is one of the clearest areas where corporate customers will pay for tools that promise immediate returns, and by tying its newest models to vulnerability detection and resilience, OpenAI is pushing deeper into the corporate stack.

For Europe, the strategic question cuts both ways. The region’s companies are clearly willing to adopt frontier AI when it improves defenses and reduces operational risk. But every new deployment also deepens dependence on a U.S. provider at a time when European policymakers continue to talk about digital sovereignty and strategic autonomy. The result is a familiar European compromise: embrace the best available technology now, while trying to keep enough control over its use to preserve long-term leverage.