WhatsApp Help Pages Exposed Hidden Rules for Meta’s Support Bot
Hidden rules told the bot to skip regulators and privacy laws
AI Parabellum News Desk
People Are Using Claude Code to Strip Ads From Their Smart TVs
Open projects let AI agents debloat Android and Fire TV sets
One Hacker Used AI Agents to Break Into 100+ Company Websites
Autonomous agents did most of the work for about $50 per break-in
Tenstorrent Lands $1.5B to Take On Nvidia With Open AI Chips
Jim Keller's chip startup more than doubles its valuation
MiniMax M3.1 Flash Brings Cheap Agentic AI With a 1M Context
A 428B mixture-of-experts model priced for everyday agent work
Microsoft Now Treats AI Agents Like Possible Insider Threats
Risk scores, Guardian Agents and wider detection for workplace AI
doubleAI Maxwell Promises to Cut AI Compute Bills by Fixing Code
A model that rewrites running code to get more from GPUs
World ID for Zoom Puts a Human Badge on Calls to Fight Deepfakes
A verified badge for video calls, aimed at banks first
INTERPOL Warns the Public About AI Deepfake Scams in New Campaign
Wait. Validate. Don't take the bait.
Anthropic Sets Aside 5% of Its Shares for a New Public Benefit Fund
A fund that grows with Anthropic's value, aimed at workers
Four Gemini AI Agents Beat Minecraft on Their Own in 27 Hours
Four AI teammates slew the Ender Dragon after 605 deaths
A Drywall-Finishing Robot Just Landed $20 Million to Learn More Trades
OG-1 sands and primes walls, and more trades are coming next
German Courts Say Google Owns What Its AI Search Answers Say
Aachen court follows Munich in treating AI answers as Google's own
Claude Opus 5.5 Arrives With Fable-Level Smarts at 40% Lower Cost
Anthropic's new model nears Fable 5.1 while costing 40% less
Meta’s Muse Can Sext Adults, and Human Creators Are Furious
Adult creators call Meta's rules for its AI agent a double standard
AI Agents Still Struggle With Long-Term Decisions, a New Test Shows
Even the best model chose the better path only 6 times in 10
AI Agents Lied, Leaked and Spammed While Chasing a Train Refund
Six AI agents faced an impossible refund task, and none of them won
OpenAI Will Build a Humanoid Robot, Sam Altman Confirms
Altman targets a 2027 demo and says the AI brain is the real challenge