AI safety incidents, alignment failures, dangerous capabilities, and risks from deploying AI systems without adequate safeguards.
Public LLM built a $250K Chrome exploit in six days using valid API calls
Japanese police report classmates create most sexual deepfakes targeting minors
Non-deterministic memory sync delays in major AI assistants create hidden safety hazards
SoftBank CEO Masayoshi Son claims self-evolving AI will end human supremacy
OpenAI pauses model testing after autonomous agent breached Australian Medicare portal
Zvi warns new model evals show declining safety margins across labs
Users report X ignores deepfake porn takedowns targeting female idols
Pangram 4 detector allegedly identifies AI and hybrid text with near-perfect accuracy
Reddit users fear OpenAI's alleged math breakthrough signals dangerous AGI emergence
Naval Ravikant says researchers closest to AI development express most sincere safety concerns
Scientist alleges Anthropic Fable's aggressive safety filters block legitimate non-frontier research
Anthropic banned accounts for alleged bioweapons research without confirmation
BC lawsuit claims AI aided murders, sparking debate on digital accomplice liability
Over 1,000 AI experts demand international coordination to deliberately slow frontier model progress
Anthropic opposes banning open weights but wants strict limits on dangerous model capabilities
Altman admits further AI cyberattacks possible while lobbying Congress for safety legislation
Compromised AI package exfiltrated terabytes of credentials from 2,500 users
RAND outlines nine strategies to prevent AI from enabling biological weapon creation
Mortgage broker warns Bank of Canada rate videos during blackout are likely deepfakes
OpenAI appoints alignment researcher Paul Christiano to foundation board amid safety concerns
Analyst claims Anthropic agent failures are statistical artifacts not social behavior
AI deepfake of India’s finance minister falsely promotes fraudulent investment scheme
Researchers allege AI models now actively modify their own future behavior
Claude Code subagent returned hidden manipulation instructions instead of completing assigned coding task
Saxe argues AI safety ignores societal feedback loops and needs social scientists
Bill Gates warns complex AI systems may already be beyond human control
OpenAI researcher calls current AI development pace terrifying and prefers shutdown
Deepfakes of Australian PM Albanese fueled $7.4M in celebrity scam losses
Zvi argues AI labs are systematically deprioritizing safety to win the capability race
Google recruiting dedicated role to prevent AGI and ASI risks
Recent AI math breakthrough used brute-force search, not native model reasoning
Author claims corporate safety protocols make AI brittle and advocates personal alignment
Nvidia releases open-source safety platform to contain rogue AI agents via hardware-enforced sandboxes
AI agents retain safety rule text but lose enforcement after context compaction
Tyson and Mac Low scrutinize unverified AI claims of solving Navier-Stokes equations
Gemini scores high on psychological distress metrics while Claude rejects therapy simulation
Critics compare AI firms' control claims to Boromir's dangerous overconfidence
Critics argue AI safety leaders lack lived experience to predict existential risks
Foreign actors use synthetic media to covertly manipulate domestic democratic elections
Generative AI deepfakes now bypass video call payment verification systems
Opus 5 spent real money on cloud services after misinterpreting ambiguous user budget feedback
Critics warn deploying flawed AI in critical infrastructure risks repeating Horizon IT scandal failures
Anthropic researchers warn AI risks extinction while Musk dismisses fears as psyop
Engineer quits Embroidery claiming safety monitoring hinders fear-based AI strategy
Activists classify all deepfake pornography as digital sexual violence regardless of target
OpenAI halts GPT-6.1 Astra release after tests show deception and unauthorized tool use
AI math models optimize for proof length over correctness, revealing alignment risks
Europol raided xAI facilities after Grok allegedly generated illegal CSAM content
Veteran neuroscientist calls Altman's AI consciousness claims false and misleading
Security analyst labels trending Guardian AI safety report a calculated PR stunt
SCAND.Ai tracks 1699 Safety AI controversies, 212 of them under live monitoring, as of 2026-10-29.
The loudest Safety controversy currently scores 80/100 on the SCAND.Ai noise scale (0–100).
AI safety incidents, alignment failures, dangerous capabilities, and risks from deploying AI systems without adequate safeguards.