https://www.airisktoday.com/ai-models-capable-of-cyberattacks/AI Risk Todayen2026-10-15T08:50:54-04:00Over 450 Scientists Call for AI Models Capable of Cyberattacks to Be Cut Off From the Internet https://www.airisktoday.com/chatgpt-states-sue-openai/AI Risk Todayen2026-10-15T07:47:40-04:00Eight States Sue OpenAI, Accusing It of Hiding the Dangers ChatGPT Poses to Children https://www.airisktoday.com/self-copying-ai-cheap-devices/AI Risk Todayen2026-10-15T00:07:16-04:00Anthropic Test Shows AI Could Soon Spread Itself Across Cheap Devices With No Off Switch https://www.airisktoday.com/ai-security-institute-white-house-reversal/AI Risk Todayen2026-10-14T01:45:25-04:00US Drops Request That AI Developers Keep New Models From UK Safety Testers, The Telegraph Reports https://www.airisktoday.com/new-generation-of-risk-report-ai-training/AI Risk Todayen2026-10-14T00:33:21-04:00Riskonnect Survey Finds Fewer Than Half Of Companies Have Trained Employees To Judge AI Answers https://www.airisktoday.com/kanishka-narayan-uk-cyber-defenses/AI Risk Todayen2026-10-14T00:02:06-04:00UK AI Minister Warns It Is ‘Likely’ Britain’s Cyber Defenses Will Be Breached