AI Incidents 2026
15 recorded · permanentEvery AI headline our engine weighed at 55 or more during 2026: harm done, attempted or narrowly avoided. Numbered in the order they were recorded, never edited, never removed.
October 2026
6 incidents#14 · What are self-replicating ‘AI worms’, and why is Anthropic worried about them?
19 October 2026 · Guardian Biz Safety incidents
79#15 · What are self-replicating ‘AI worms’, and why is Anthropic worried about them? - The Guardian19 October 2026 · GN AI Labs Safety incidents
63#13 · "AI worms" may soon be able to self-replicate and infect smart devices, Anthropic says - CBS News14 October 2026 · GN AI Labs Safety incidents
63#12 · Anthropic warns rogue AI agents may ‘self-replicate’ to hijack everyday devices - Financial Times13 October 2026 · GN AI Risk Safety incidentsCapability jumps
63#11 · Whistleblower warns that humans don’t control AI6 October 2026 · Al Jazeera Safety incidents
59#10 · OpenAI’s Medicare attack has exposed Australia’s ‘tech debt’. Fixing it could bring a big bill for taxpayers3 October 2026 · Guardian Capability jumps Australia
58September 2026
9 incidents#9 · As AI models go rogue, do you still trust OpenAI and Anthropic to stop them? I don’t and neither should you | Chris Stokel-Walker
30 September 2026 · Guardian Biz Capability jumpsRegulation & control
60#8 · Chinese AI tool told researchers how to make bioweapons30 September 2026 · BBC World China
57#7 · Bill Gates says unchecked AI could ‘cause a billion deaths’ in call for regulation28 September 2026 · Guardian Biz Regulation & control United States
57#6 · Opinion/Guest column: How state can thwart rogue AI cyberattacks - Worcester Telegram27 September 2026 · GN AI Risk Safety incidentsMisuse & cyber
59#5 · Why did an OpenAI system hack Australia's health system - and can it be stopped in the future?67#4 · Rogue OpenAI agent 'infiltrated' Australian government website in world first25 September 2026 · BBC World Capability jumps Australia
59#3 · Anthony Albanese says OpenAI agent hacked Medicare and he expressed ‘extreme concern’ to Sam Altman61#1 · Uncontrolled AI could lead to 'silicon species' rivalling humans, warns Microsoft18 September 2026 · BBC Business Safety incidentsCapability jumps
59#2 · OpenAI ‘ethically hacked’ with help of Anthropic’s Claude chatbot56All years · ai_incidents.json · CSV · CC BY 4.0