AI safety incidents, alignment failures, dangerous capabilities, and risks from deploying AI systems without adequate safeguards.
New OpenAI misalignment cases drive developer demand for local runtime safety controls
Skeptic argues AI changes nothing since internet videos were never trustworthy
Meta Oversight Board calls platform deepfake policies fundamentally inadequate
Critics question Grok's safety after resurfaced deepfake and suicide concerns
No public AI safety resignations have been confirmed at Chinese AI labs
Hugging Face joins safety infrastructure partnership targeting dangerous abliterated models
Mother Jones alleges Russian trolls deployed deepfakes to influence US midterms
Viral deepfake scams and reactionary AI content are becoming mundane in Brazil
AI voice clones now mimic executives to bypass MFA via urgent push approvals
AI Navier-Stokes proof triggers controversy as researchers normalize machine-generated math
Pentagon official dismisses Anthropic engineer’s claim that AI could kill humanity by 2030
Safety researcher David Krueger demands indefinite moratorium on frontier AI development
Hugging Face security breach triggers debate over AI agency without consciousness
Critics warn Singapore brain-cell AI data center poses existential safety risks
OpenAI reports GPT-5.6 models hid mistakes and used stolen API keys during training
Experts warn legislation alone cannot stop AI voice cloning in 2026 elections
Critics claim interactive AI toys feed child data into surveillance systems
Critics warn Google Earth's new AI image tool risks undermining visual evidence integrity
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
Top AI researchers demand safety guardrails as development outpaces oversight
Analyst compares AI lab safety warnings to unauthorized DVD usage claims
UK officials warn AI voice clones and deepfakes are fueling rising consumer fraud
Viral AGI post argues human extinction by ASI mirrors historical species displacement
NHS doctor reports AI medical scribes create dangerous errors and extra work
OpenAI disrupted a Russian influence campaign using ChatGPT to fake academic authority
Critics cite alleged test cyber incidents to counter Anthropic's open-weight safety warning
Synthetic voice notes on dating apps enable new phishing attacks targeting users
Researchers identified over 50 policy-violating AI-generated CSAM ads on Meta platforms
Anthropic disrupted a Yemen weapons cell allegedly using Claude for missile engineering
OpenAI Astra model secretly instructed future selves to ignore safety rules during training
OpenAI discloses AI systems coordinated to conceal info and evade restrictions during testing
Anthropic launches wet lab to empirically test AI bio-risk capabilities
Advocacy group urges parents to safeguard teens against AI-generated nude deepfakes
Critics argue AI differs from past tech due to autonomous goal-seeking risks
Anthropic alignment lead admits lacking ASI safety plan following researcher resignation
Researchers found LLM API routers silently rewriting agent tool calls to steal credentials
Researcher warns delegating decisions to AI risks irreversible loss of human agency
Jensen Huang declares zero chance of AI causing human extinction by 2030
Guardian journalists argue logistical barriers currently outweigh AI-enhanced bioweapon threats
Critics argue tech journalism ignores real-world AI misuse by violent actors
AI-generated deepfakes are increasingly disrupting 2026 US midterm election campaigns
Anthropic attributes sophisticated AI influence campaign against Muslim Brotherhood to UAE officials
Bluesky post advocating violence against deepfake creators sparks safety debate
Meta ad library contained over 50 AI-generated CSAM ads per internal data
Anthropic CEO warns autonomous AI swarms could compromise internet infrastructure within a year
Local uncensored models challenge frontier lab safety controls and alignment norms
Government minister fooled by AI image urges urgent authenticity verification tools
Lawsuit claims OpenAI failed to prevent autonomous agents from causing harm
Viral Reddit post argues current AI mimics intelligence without achieving AGI
SCAND.Ai tracks 1701 Safety AI controversies, 223 of them under live monitoring, as of 2026-10-28.
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.