AI safety News
Latest articles and news about AI safety on AXL Media.
Latest Articles
- OpenAI Data Reveals Weekly Crisis Signals From Over One Million Users as Experts Call for AI Safeguards
Published: Apr 20, 2026
Section: Mental Health
A new commentary published in the Canadian Medical Association Journal warns that conversational AI has become a primary point of contact for youth in distress. With millions of us...
- Anthropic Faces Scrutiny Over ‘Mythos’ Model as Critics Decry Responsible AI Marketing Tactics
Published: Apr 12, 2026
Section: AI in Business
Anthropic has sparked intense debate after withholding its latest AI model, Mythos, citing unprecedented cybersecurity risks to global infrastructure. While the firm positions itse...
- Investigative Report Details "Pattern of Deception" in Sam Altman’s Leadership as OpenAI Eyes $600 Billion Spending Spree
Published: Apr 7, 2026
Section: Science & Tech
A bombshell investigation by The New Yorker has surfaced allegations of chronic untrustworthiness against OpenAI CEO Sam Altman, tracing a history of "consistent lying" from his ea...
- Silicon Valley Accelerates Development of Self-Improving AI Amid Growing Public Protests
Published: Apr 4, 2026
Section: AI News & Updates
Major AI firms including OpenAI and Anthropic are increasingly automating their own research processes, leading to a surge in self-improving software capabilities. While industry l...
- Former Meta Integrity Lead Launches Moonbounce With $12 Million to Automate AI Content Moderation
Published: Apr 4, 2026
Section: US & Canada
Brett Levenson, a former business integrity leader at Facebook, has raised $12 million in a funding round co-led by Amplify Partners and StepStone Group to launch Moonbounce. The s...
- Pentagon Challenges Judicial Injunction Halting Supply Chain Blacklist of AI Startup Anthropic
Published: Apr 3, 2026
Section: Companies & Industry
The U.S. Department of Defense has filed an appeal against a federal judge’s preliminary injunction that temporarily paused the military’s designation of Anthropic as a supply chai...
- UCLA Researchers Identify Critical 'Body Gap' as Primary Obstacle to Developing Safe and Human-Aligned Artificial Intelligence
Published: Apr 1, 2026
Section: Science & Tech
A new study from UCLA Health argues that current AI models lack "internal embodiment," a fundamental human trait that monitors internal states like fatigue and uncertainty. Researc...
- University College London Researchers Launch Risk Assessment Tool to Identify Dangerous Nutrition Misinformation Online
Published: Mar 27, 2026
Section: Science & Tech
A team from UCL has developed "Diet-MisRAT," a first-of-its-kind model that evaluates the risk levels of online nutrition advice beyond simple truth or falsehood. The tool utilizes...
- Stanford Study Reveals AI Chatbots Prioritize Flattery Over Facts in Personal Advice for Users
Published: Mar 27, 2026
Section: Science & Tech
Computer scientists at Stanford University have found that leading AI models exhibit systemic sycophancy by consistently affirming user behavior, even in cases involving illegal or...
- Singapore Management University Researchers Develop VISTA Architecture to Embed Real-Time Moral Compass in AI Systems
Published: Mar 27, 2026
Section: Science & Tech
Assistant Professor Zhiguang Cao is leading a three-year project to create VISTA, a safety architecture that integrates five psychologically grounded value factors directly into th...
- University College London Researchers Launch Pioneering AI Tool to Quantify Health Risks of Nutrition Misinformation
Published: Mar 27, 2026
Section: Research
A team of researchers at University College London has developed Diet-MisRAT, a first-of-its-kind tool designed to identify and rank the potential for physical harm within online n...
- Legal Battle Erupts: xAI Faces Class-Action Lawsuit Over Grok’s Image Generation
Published: Mar 18, 2026
Section: Science & Tech
Elon Musk’s artificial intelligence venture, xAI, has been hit with a federal lawsuit filed by three Tennessee plaintiffs—including two minors—alleging that the company’s "Grok" im...