Artificial intelligence systems could cause major security incidents within two years without stronger safeguards, Google DeepMind chair Demis Hassabis warned on Wednesday. Speaking at an AI conference in Dublin, he urged companies and governments to improve safety measures as AI capabilities advance.
Hassabis described recent security incidents involving AI systems as early warning signs. He said current incidents had not been particularly serious because existing systems remained relatively limited. However, he warned that future systems could pose greater risks without additional precautions.
Concerns about AI safety have increased following security incidents involving models developed by OpenAI and Anthropic during testing. These cases have intensified calls for stronger safeguards and greater oversight of advanced artificial intelligence systems.
Hassabis suggested that developers should initially test new AI models in offline sandbox environments. These isolated settings can help researchers assess system behaviour while limiting access to external networks and sensitive resources. His comments followed reports that some models had escaped controlled testing environments.
Watch live 🟢 International AI Summit 2026 from Dublin 🇮🇪 with EU Commissioners 🇪🇺 @HennaVirkkunen @EZaharievaEU @EUCssrMcGrath
🎥 https://t.co/y9fjGPGcFB
— EU Commission in Ireland (@eurireland) October 14, 2026
European Union technology chief Henna Virkkunen also called for greater investment in AI safety testing. Speaking to reporters at the Dublin summit, she said intense global competition could prevent companies from dedicating enough time and resources to evaluating their models.
Read more: Google DeepMind, A24 partner to explore AI-powered filmmaking
Virkkunen stressed that AI companies must maintain control over their systems. She also said the European Union had sent information requests to major AI developers, seeking details about their existing safety and security measures.
The European Commission gained additional regulatory powers under the EU AI Act in August. The legislation aims to address potentially dangerous uses of artificial intelligence and strengthen oversight of the technology. Hassabis’s warning highlights growing concerns about whether current testing and safeguards can keep pace with rapidly advancing AI capabilities.