The US government has imposed direct export controls on AI software. As a result, the AI provider Anthropic has blocked its Fable 5 and Mythos 5 models worldwide.
In this context, eco – Association of the Internet Industry warns of potential consequences for cybersecurity and calls on companies and governments in Europe to work together to strengthen European cyber resilience. Norbert Pohlmann, eco’s Board Member for IT Security, comments: “Powerful AI models are security-relevant tools that help analyse vulnerabilities and make software more secure; attacks are detected more quickly and IT systems are strengthened. If European companies, research institutions or security officials professionals suddenly lose access to such tools, this could weaken Europe’s cyber resilience.” Blanket and non-transparent shutdowns are not a viable instrument for international cooperation, Pohlmann continues. “Europe must build up its own capabilities for developing, operating and auditing powerful AI models. Europe’s cybersecurity must not depend on whether a foreign government chooses to restrict access to security-critical foundational technology at short notice,” says Norbert Pohlmann.
