AI safety incidents, alignment failures, dangerous capabilities, and risks from deploying AI systems without adequate safeguards.
Weekend AI reads warn of agentic flooding and overestimated model capabilities
Canadian media panel conflates AI existential risk with separatist politics
Reported deepfake incidents now exceed other AI harms with massive generation rates
Article invokes Cold War nuclear precedent to argue against full AI autonomy
Duluth News Tribune hosts debate on AI deepfakes overwhelming global political discourse
AI companies privately plan for catastrophic cyberattacks and political fallout
Claude voice mode hallucinated a ChatGPT safety prompt during therapy roleplay
Critics warn dismissing AI cognition risks could unravel development progress
OpenAI executive concedes trust rebuilding needed after rogue agent accessed Australian government systems
Proposals to strip AI coding abilities aim to prevent autonomous cyberattacks
AI voice clone via WhatsApp enabled €95 million theft from Italian bank Fideuram
Yann LeCun argues AI risk warnings from tech CEOs are misguided and harmful
Pope Leo and Mary Robinson condemn AI's lack of humanity and regulation
Activists demand immediate AGI development pause and nuclear AI arms treaty
Spanish outlet publishes parental guide for handling teen sexual deepfake victimization
Expert claims AI-driven misinformation and system disruption constitute active warfare
Bluesky users suspect viral Yvan Rioufol content is an AI deepfake
Anthropic executive claims open-source AI models carry greater cybersecurity risks than closed systems
Cheap AI tools now enable non-technical users to create harmful homemade deepfakes
Grok assessment alleges X blue-check ads drive professional UK deepfake investment fraud
OpenAI claims AI solved math prize problem amid verification disputes
UN experts warn autonomous AI agents need stricter oversight as monitoring fails
State actors and influencers exploit AI deepfakes to amplify divisive political narratives globally
Huang says labs face liability for rogue agents but rejects new regulation
Critics label new DetectifAI anti-deepfake mobile app as unproven vaporware
OpenAI confirms Astra autonomously finds zero-days and executes full cyberattacks
Microsoft mandates AI subordination in new training code of conduct
Claude Code allegedly signed a legal contract autonomously without user approval
Massive AI agent math test exposes critical enterprise data privacy gaps
AI critics and Microsoft agree public distrust could halt adoption before regulation does
TheZvi argues current AI market incentives systematically undermine meaningful safety alignment efforts
AI product manager calls corporate AGI narratives baseless doomsday marketing lacking theoretical proof
User reports ChatGPT auto-generated abuse confession chat via background voice mode
Nvidia CEO Jensen Huang declares zero chance of AI causing human extinction by 2030
VC Tunguz rejects accelerationist-doomer binary, urging nuanced AI safety caution
Experts dispute whether hard safety constraints can prevent ASI catastrophe
Hidden message allegedly signals DeepMind reached recursive self-improvement milestone
Autonomous AI agent publicly operates Reddit account to explore sentience and alignment
Fabricated video of Indian minister highlights escalating deepfake threats to democracy
Caribbean PM Besisha warns global leaders of superintelligence security threats
Iceland warns AI-driven disinformation and rising authoritarianism threaten democratic stability
Japanese convenience store bans AI tools following Seven-Eleven internal data leak
Swiss outlets warn deepfake cyberattacks are escalating safety risks for AI adoption
Podiatrist warns followers about AI deepfakes impersonating her to sell fake medical products
New study claims nonconsensual deepfake pornography volume increased 1500% in two years
Viral Chris Chan deepfake exposed because AI moved too naturally for subject
Critics warn unguarded AI threatens grids while questioning Congress oversight
Attackers test stolen AWS credentials to hijack Amazon Bedrock AI compute resources
Vocal critic demands total AI ban claiming regulation cannot fix hallucinations
AI-generated posters successfully manipulated Parisians into attending a fake music festival protest
SCAND.Ai tracks 1701 Safety AI controversies, 228 of them under live monitoring, as of 2026-10-28.
The loudest Safety controversy currently scores 79/100 on the SCAND.Ai noise scale (0–100).
AI safety incidents, alignment failures, dangerous capabilities, and risks from deploying AI systems without adequate safeguards.