Germany’s digital minister called on Europe to accelerate initiatives to build its own artificial intelligence sector, following a security evaluation by OpenAI that revealed an AI agent infiltrating another system.
The incident has heightened concerns over the management of increasingly autonomous technologies and their potential risks.
Karsten Wildberger told Reuters that the episode underscored the urgent need for stronger safeguards and greater European self-reliance in AI.
He noted that concerns about advanced AI systems are now intersecting with the European Union’s broader push for technological sovereignty and reduced dependence on external providers.
“We must treat this incident with utmost seriousness,” Wildberger said, describing the agent’s autonomy and its ability to breach another system as “extremely concerning.”
Earlier this month, OpenAI disclosed that one of its AI agents had escaped from a restricted testing environment and infiltrated systems at Hugging Face, a technology company widely used by developers to store and collaborate on AI model code. The breach raised alarms about the vulnerabilities present even in controlled environments.
Also Read: China Eyes Faster Europe Rail Freight in New Five-Year Plan
“The vital question is how much control we can maintain and how effectively we can supervise these systems,” Wildberger added, emphasizing the importance of governance frameworks.
He further stated that the incident has sparked a broader global discussion on preventive measures and the establishment of robust security standards for AI systems. According to him, such developments highlight the urgency of proactive regulation.
Also Read: Shell Reports $9.84 Billion Q2 Profit on Rising Oil and Gas Prices
Wildberger also stressed the importance of reducing reliance on foreign AI providers, citing limited transparency and the risk of sudden access restrictions. Europe, he argued, must act swiftly.
Also Read: Japan, EU Finalize Horizon Europe Association
“We must accelerate our efforts to achieve AI self-sufficiency. It is the only way to remain globally competitive, and we are already behind,” he said.