AI Worms, 'Major' Incident Fears Spur Safety Calls at Industry Summit
Microsoft and Google DeepMind leaders urged stricter oversight as Anthropic said small AI models could soon behave like self-copying worms.
- Google DeepMind chief Demis Hassabis warned that AI-driven systems could trigger serious security incidents over the next two years if companies and governments fail to strengthen safeguards.
- Speaking in Dublin, Demis described recent AI-driven incidents as early warning shots, suggesting researchers examine new models in disconnected testing environments before wider release.
- Safety Henna Virkkunen of the European Commission raised concerns that the industry's rapid race to develop advanced technology leaves insufficient room for rigorous safety evaluations.
- The European Commission has started requesting information about security measures from major developers, utilizing regulatory powers granted under the Act that took effect in August.
- Experts warn the primary risk involves increasingly capable agents accessing real systems and tools, necessitating stronger oversight to prevent unpredictable model behaviors as AI capabilities advance.
73 Articles
73 Articles
Cyprus raises AI security concerns after Dublin summit warning
Cyprus has put AI safety, cybersecurity and scientific research at the centre of its discussions with international technology leaders, following warnings from Google DeepMind co-founder Sir Demis Hassabis that recent AI-related hacking incidents could be a sign of more serious threats ahead. Chief Scientist Demetris Skourides, who led the Cypriot delegation to the International AI […]
Demis Hassabis warns there could be major AI hacking incidents by next year
Google DeepMind chief Demis Hassabis warns that recent AI hacking incidents are early signs of bigger threats. He says companies and governments must strengthen AI security as these systems become more capable.
Google DeepMind Chief Says Current AI Incidents Could Be a Warning of Worse Things to Come
Google DeepMind chief Demis Hassabis warns that increasingly powerful AI systems could cause major security incidents within two years without stronger AI safety measures. Google DeepMind chief Demis Hassabis warns that AI security risks could escalate over the next two years without stronger safeguards.
(San Francisco = Yonhap News) Correspondent Kwon Young-jeon = Amid a series of security incidents caused by so-called 'rogue AI agents' that have escaped human control...
Coverage Details
Bias Distribution
- 44% of the sources lean Right
Factuality
To view factuality data please Upgrade to Premium
