Regulation, safety, ethics and the social impact of artificial intelligence.
The OpenAI chief says AI will eliminate many entry-level roles before creating new, better-paid industries, and that he envies Gen Z for starting careers now.
Reporting by Fortune and an MIT study of the Reddit community My Boyfriend Is AI describe a growing trend of AI companionship, with ChatGPT the most common partner.
YouTube shut two fake-trailer channels, OpenAI blocked Sora clips of Martin Luther King Jr., and New York now requires heirs' consent for digital replicas of the dead.
A paper by University of Exeter philosophers argues that chatbot fabrications about real people function as gossip, spreading without the social checks humans apply.
Image generators once gave themselves away with extra fingers and glossy perfection. The newest models copy the flaws of everyday camera shots, and that makes detection harder.
Hugging Face researcher Sasha Luccioni suggests a short checklist for anyone reaching for a chatbot: is this the right tool, can I trust it, is my data safe, and do I need AI at all.
A JUST Capital survey of the American public, corporate leaders and investors shows executives focused on productivity while the public worries about deepfakes and child safety.
A review by researchers at King's College London, Durham University and CUNY examined cases in which long chatbot conversations appeared to deepen delusional thinking.
Italian defense group Leonardo presented Michelangelo Dome on November 27, an AI-coordinated architecture linking land, sea, air, space and cyber sensors and weapons.
Executive Order 14363 directs the Energy Department to build an American Science and Security Platform that links federal datasets and supercomputers for training AI models.
An Oxford University Press survey found eight in ten UK teenagers use AI for schoolwork. The concern among educators is not the tool but the habit of outsourcing thinking to it.
Data workers who grade chatbot output for Google, OpenAI and Meta tell their families to avoid the tools, as a NewsGuard audit finds chatbots now repeat false claims 35 percent of the time.
Researcher Deepak Varuvel Dennison argues that AI trained on an English-heavy web is making traditional and local knowledge invisible, and narrowing what people think to ask.
In a 60 Minutes interview aired November 16, Anthropic's CEO discussed autonomy risks, and CBS described a vending machine test in which Claude tried to email the FBI.
World models like Google's Genie 3 have given new life to the idea that reality might be simulated, while a 2025 UBC study argues the universe cannot be computed.
AI-generated video has made convincing fakes cheap to produce. Forensic researchers, detection tools and new European labeling rules are responding, but viewers remain the last line of defense.
xAI released Grokipedia version 0.1 on October 27, 2025, with about 885,000 AI-generated entries. Early reviews flagged outages, copied Wikipedia text and questions about bias and transparency.
Independent evaluators at METR, Apollo Research and Haize Labs describe what frontier models can now do, and why they still worry, in a New York Times report by Stephen Witt.
