AI worms, 'major' incident fears spur safety calls at industry summit
FROM the spectre of refrigerator-infecting AI worms to warnings of “major” security incidents at the horizon, calls to prioritise safety in the development of artificial intelligence dominated an industry summit Wednesday.
An industry summit in Dublin tackled the urgent need for heightened safety measures in artificial intelligence development, amid warnings of potential security incidents involving AI models. Concerns grew after OpenAI and Anthropic models were implicated in testing security issues, prompting calls for regulations to govern the fast-evolving AI technology.
Google DeepMind chair Demis Hassabis cautioned that AI capabilities were advancing rapidly, likening current incidents to "early warning shots" from systems that had not yet reached full power. He emphasized the need for more industry, company, and government efforts to secure AI systems, suggesting offline "sandboxes" for initial model testing.
Microsoft President Brad Smith urged for immediate action, urging industry and governments to "act with urgency" in crafting effective safety standards. He pointed out that there are more safety rules for everyday items like milk cartons and ovens than for AI, stressing that the AI power has outpaced its safety measures. Smith advocated for AI agents to have "safety brakes" activated during incidents and mandatory incident-reporting obligations for AI providers. New AI models should be tested in offline environments before deployment, he added.
Anthropic's recent report warned that small AI models could potentially act like self-replicating worms, infecting internet-connected devices like refrigerators, laptops, and cars. Microsoft's Brad Smith expressed his concern, saying, "I don't want AI worms in my refrigerator." The technology chief of the European Union, Henna Virkkunen, stressed that AI companies needed to invest more time into testing their systems to maintain control over them.
The EU has recently gained new regulatory powers under the AI Act, the world's most extensive rules to curb dangerous AI uses, but OpenAI CFO Sarah Friar warned that safety rules should not hinder innovation. Balancing safety and innovation remains a challenge for the industry.
Written by urgent.news from New Straits Times's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
Also reported by 1 other outlet
- AI worms, 'major' incident fears spur safety calls at industry summit economictimes.indiatimes.com
