EU confirms OpenAI offers access to cybersecurity model, Anthropic lags behind

OpenAI is handing EU cybersecurity officials the keys to its latest AI defense tool. Anthropic, its closest competitor, hasn’t made a similar offer, and European regulators have taken notice.

The European Commission confirmed that OpenAI committed on May 11, 2026, to provide access to GPT-5.5-Cyber, a specialized cybersecurity model, for evaluation and potential deployment by EU cyber defenders. Anthropic, meanwhile, has not extended access to its own advanced model, Mythos, creating a visible gap in how the two AI heavyweights are engaging with one of the world’s most powerful regulatory bodies.

What OpenAI is actually offering

GPT-5.5-Cyber is not just a marketing exercise with a security-themed name slapped on it. The model recently demonstrated its capabilities by completing a 32-step cyberattack simulation in 12 hours, suggesting it can model and anticipate complex attack chains that would take human teams considerably longer to map out.

The EU’s interest in this kind of tool makes sense when you consider the scale of the problem it’s trying to solve. In the crypto industry alone, cyber threat-related costs exceeded $1.7 billion in 2025. This evaluation fits squarely within the framework of the AI Act, which mandates enhanced oversight of high-risk AI systems, particularly those deployed in critical areas like cybersecurity.

Anthropic’s conspicuous silence

As of May 13, 2026, Anthropic has not offered access to Mythos for similar EU review. The company, which has built its brand around AI safety and responsible development, finds itself in an awkward position.

Why crypto should be paying attention

The intersection of AI cybersecurity models and blockchain infrastructure is no longer theoretical. AI tools are increasingly deployed for threat detection across decentralized protocols, exchange platforms, and DeFi applications. When the EU evaluates a model like GPT-5.5-Cyber, it’s also implicitly setting standards for what AI-powered security looks like in regulated environments, including crypto markets.

The $1.7 billion in crypto-related cyber losses from 2025 represents both a problem and an opportunity. Protocols and exchanges that can point to AI-enhanced security measures may find themselves on better footing with regulators who are simultaneously tightening rules around digital asset platforms.

OpenAI’s proactive engagement could also influence investor sentiment toward AI-integrated security solutions in the crypto space. Analysts believe that OpenAI’s collaboration could pave the way for future integrations between AI and crypto, potentially influencing investor sentiment toward AI-focused digital assets as market players await further developments.