Topic
AI Safety
Who decides what these systems are allowed to do.
Latest stories
- OpenAI flags new concerning AI behaviour, to track model misalignment regularly (2026-09-17)
- AI pioneer Yoshua Bengio warns humanity is losing control of AI technology — The co-inventor of backpropagation echoes an emerging consensus among safety researchers that governance frameworks are urgently needed. (2026-09-17)
- Microsoft AI chief flags risks in Anthropic's development model — Public disagreement between frontier labs over safety strategy reflects deepening fault lines on alignment. (2026-09-17)
- OpenAI publishes six cases of model misbehaviour, commits to systematic reporting — The company moves toward formal transparency on safety incidents, though none of the disclosed cases caused material harm. (2026-09-17)
- Divisions emerge in tech industry over coordinated AI slowdown proposals — Industry consensus fractures as some leaders push back against calls for development restrictions. (2026-09-16)
- Anthropic and OpenAI to embed independent safety evaluators in their labs — Companies seek outside oversight to manage risks, but researchers doubt independence without regulatory teeth. (2026-09-16)
- US White House opposes AI oversight as safety concerns mount — Political resistance to regulation is hardening even as incidents of model misbehaviour multiply. (2026-09-16)
- UN chief calls for global cooperation on AI safety standards — The secretary-general warns that competitive pressure risks eroding safety practices across nations. (2026-09-16)
- New report warns AI data centre e-waste could reach 23 million shipping containers by 2050 — The environmental cost of training and running frontier models is far larger than previously calculated. (2026-09-16)
- Questions linger over mandatory incident reporting for AI companies — As AI risks grow, regulators have yet to establish whether companies must publicly disclose dangerous incidents. (2026-09-16)
- Trump adviser Sacks says AI safety concerns driven by fear-mongering — A senior Trump official frames mainstream AI safety discourse as a tactic to slow American innovation. (2026-09-16)
- AI labs seek in-house auditors to manage rogue agents — Internal oversight may miss the obvious: restricting what agents can do in the first place. (2026-09-16)
- Al Gore prioritises AI safety concerns above data centre environmental impact — Industry warnings about AI development matter more than computational power's energy use. (2026-09-16)
- DeepMind cofounder warns AI capabilities may outpace safety controls — Increasingly capable AI agents pose risks if safeguards cannot keep pace with recursive self-improvement. (2026-09-16)
- AI-powered dating app scams targeting users with synthetic voices — Fraudsters are automating romance schemes with generative audio, making catfishing scalable and harder to detect. (2026-09-16)
- Expert suggests superintelligent AI may not act maliciously toward humans — The risk lies not in AI intent but in human misalignment with machine objectives. (2026-09-16)
- Microsoft AI chief raises safety concerns over Anthropic's Claude consciousness approach — Debate over whether training AI systems on consciousness concepts creates uncontrollable entities shifts from academic to boardroom. (2026-09-16)
- Sam Altman uses Hugging Face hack to promote OpenAI's cyber-defence solutions — OpenAI's CEO weaponised a security breach to sell enterprise customers on AI-powered defence tools. (2026-09-16)
- Leading AI companies acknowledge safety concerns but face obstacles to coordination — Shared commitment to responsible development remains theoretical without mechanisms to enforce it across competing firms. (2026-09-16)
- Spain's data watchdog reports first personal data breach by AI agent — Autonomous systems are now directly implicated in cyberattacks, marking a shift in the threat landscape. (2026-09-16)
- EU chief von der Leyen to unveil curbs on social media and AI for under-15s — Europe tightens restrictions on tech access for minors ahead of frontier-model proliferation. (2026-09-16)
- Zuckerberg says competition and liability ensure AI safety without coordination — Meta's CEO parts ways with peers who advocate slowing development, arguing market forces suffice. (2026-09-16)
- Jensen Huang, Sam Altman and Dario Amodei advocate for AI safety and measured development — Three of AI's most influential leaders are urging the industry to prioritise safety and transparency over velocity. (2026-09-16)
- Mark Zuckerberg urges labs to train models safely and maintain power balance — Meta's chief frames AI responsibility as a competitive necessity, not just ethics. (2026-09-16)
- Nobel laureate Maria Ressa warns of AI threatening human agency — A UN panel co-chair argues rapid AI deployment risks eroding human autonomy and control over critical decisions. (2026-09-15)
- Agility's humanoid robot squats to avoid injuring human coworkers — A safety mechanism that lets robots work alongside humans without barriers signals progress in collaborative robotics. (2026-09-15)
- Anthropic's Amodei urges AI companies to set safety standards before regulators intervene — Self-imposed standards may forestall stricter government mandates, but the industry remains fragmented on safety priorities. (2026-09-15)
- AI industry roundtable examines whether advanced AI could destroy humanity — Lab employees are articulating existential risk seriously, forcing the sector to address whether the concern is substantive or promotional. (2026-09-15)
- AI Contact Hotline created for agents to report misbehaviour — A new mechanism to surface misconduct by AI systems themselves, if they choose to use it. (2026-09-15)
- AI companies must detect dangerous model use before harm occurs — Determining misuse from individual interactions is difficult, but prevention demands early action. (2026-09-15)