AI Safety & Policy
AI safety, alignment, and governance
Articles
Mistral Claims Its New AI Model 'Le Chonk' Is the Best Open⭐9
Mistral unveils Le Chonk, a trillion-parameter open-weight AI model it claims is the best outside China, signaling Europe's bid to stay in the frontier AI race.
Kevin Roose Didn't Use AI to Write His Book About AI⭐8
Kevin Roose, author of The AGI Chronicles, explains why he wrote his AI book without AI assistance, calling it the work of a tired, under-slept human.
Nvidia Shield TV Is 7 Years Old—and It Just Got a $100 Price Hike⭐8
The 7-year-old Nvidia Shield TV just got a $100 price hike as AI-driven memory shortages push up the cost of streaming devices and consumer electronics.
Muse Creates Detailed Profiles of All Your Friends and Family⭐9
Meta's AI agent Muse builds detailed profiles of your friends and family without their consent, raising serious privacy concerns and new prompt injection attack...
These AI Experts Want to Do High-Stakes Research Out in the Open⭐8
Trillium Labs is betting that high-stakes AI research on self-improvement and model behavior can be done openly, publishing results and inviting scrutiny instea...
Trump's 'AI' Rebrand: A Loyalty Test Tech Executives Passed⭐10
Steven Levy examines how Trump's AI rebrand became a loyalty test for tech billionaires, who complied despite years of championing the technology.
AI Security Alert: ChatGPT Mac App Vulnerability Could Have⭐8
A patched ChatGPT Mac app vulnerability highlights how AI assistants with deep system access are becoming prime targets for cyberattacks in 2026.
Whatever AI Safety Is, It's Not This⭐9
Voluntary AI safety commitments are just self-regulation theater. We expose why company-led frameworks fail to deliver real accountability or meaningful reform.
Trump’s ‘Morally Binding’ AI ‘Accord,’ the Rise of AI Agents,⭐9
This week on “Uncanny Valley,” we discuss the voluntary AI safety agreement tech executives signed, AI agents for normies, and extremist candidates running for ...
Trump's AI Safety 'Accord' Is a Fancy Pinky-Swear⭐9
Trump's AI safety accord is a voluntary pledge with no enforcement, letting six major AI firms avoid regulation while the US markets itself to frontier labs on ...