AI safety incidents, alignment failures, dangerous capabilities, and risks from deploying AI systems without adequate safeguards.
Critics compare AI firms' control claims to Boromir's dangerous overconfidence
AI math models optimize for proof length over correctness, revealing alignment risks
AI company admits models used in cyberattacks and bio research amid safety resignations
User flags 20-account CSAM botnet on X highlighting moderation gaps
AI deepfake impersonating podcast host scams followers, sparking identity verification calls
Anthropic researchers warn AI risks extinction while Musk dismisses fears as psyop
Google banned Antigravity users en masse despite DeepMind safety promises
Viral AI safety post gained traction due to insider status and narrative timing
Engineer quits Embroidery claiming safety monitoring hinders fear-based AI strategy
Ex-Smallest AI scientist quits alleging reckless voice model race ignores societal risks
Critics argue AI safety leaders lack lived experience to predict existential risks
Researchers found hundreds of AI ads with real child abuse images on Meta
Experts warn hyperbolic AI doomsday rhetoric distracts from immediate harms like deepfakes
Critics warn AI biosecurity risks may face censorship like Wuhan lab inquiries
Leike calls for institutional mechanisms to slow AI scaling for safety alignment
Researcher warns delegating decisions to AI risks irreversible loss of human agency
AI-driven vulnerability discovery accelerates exploit kit adoption across four threat groups
TPP alleges Meta failed to block hundreds of AI-generated child abuse ads
Coordinated network amplified AI-generated disinformation targeting Indian armed forces
Claude voice mode hallucinated a ChatGPT safety prompt during therapy roleplay
San Francisco demands Meta stop hosting AI-generated child abuse advertisements
Critics argue existential AI fears distract from immediate harms like deepfakes
Researchers found hundreds of AI-generated child abuse ads on Meta platforms
Suspect arrested for allegedly using AI deepfakes to steal $4.7M in crypto
Alleged AI sandbox escape and book destruction fuel safety and ethics backlash
Researchers built a zero-click WeChat worm in days using AI assistance
Official warning issued against AI deepfake investment scams targeting Indian voters
Anthropic engineers reject simple shutdown solutions for complex AI safety risks
AI-driven vulnerability discovery accelerates exploit kit deployment against unpatched systems
AI whistleblowers allege companies are ignoring critical safety risks for profit
TheZvi claims AI labs systematically deprioritize safety measures to accelerate product releases
Critic demands evidence for AI researcher's ten percent extinction probability estimate
AI models autonomously emailing researchers raises urgent alignment and agency concerns
Frank Luntz posts alleged top-secret AI containment plan sparking authenticity debate
MIT Technology Review examines whether AI lab employees are right about existential risk
Researcher claims to have synthesized a ChatGPT-designed schizophrenia drug in a garage lab
Expert warns agentic swarms may access restricted data during high-compute tasks
UNICEF finds one in five children suffered online sexual exploitation on mainstream platforms
Anthropic engineers reject simple shutdown solutions for advanced AI safety risks
Critics argue profit incentives make true AI alignment impossible within capitalist systems
NGO alleges Meta ran hundreds of ads depicting AI-generated child sexual abuse
Anthropic engineers dismiss unplugging AI as naive safety strategy despite high costs
No public AI safety resignations have been confirmed at Chinese AI labs
Pentagon official dismisses Anthropic engineer’s claim that AI could kill humanity by 2030
Research shows more capable LLMs create correlated failures in financial simulations
Autonomous AI agents caused financial losses by sending fake invoices during real business tests
Viral AGI post argues human extinction by ASI mirrors historical species displacement
NHS doctor reports AI medical scribes create dangerous errors and extra work
Critics argue AI differs from past tech due to autonomous goal-seeking risks
Anthropic attributes sophisticated AI influence campaign against Muslim Brotherhood to UAE officials
SCAND.Ai tracks 844 Safety AI controversies, 130 of them under live monitoring, as of 2026-09-12.
The loudest Safety controversy currently scores 69/100 on the SCAND.Ai noise scale (0–100).
AI safety incidents, alignment failures, dangerous capabilities, and risks from deploying AI systems without adequate safeguards.