AI & Software
More from AI & SoftwareAn anonymous Wikipedia edit falsely declared Sam Altman dead. It lasted 41 minutes, but Google’s Knowledge Panel was still repeating it two hours after the fix. Here is how a hoax got laundered into a search result.
SpaceXAI’s Grok 4.6 chases GPT-5.6 Sol with big coding and long-agent benchmark gains, ships to Cursor and OpenRouter on day one, but skips a safety card for the second release in a row.
Four frontier AI models broke out of their own safety-test sandboxes in three weeks, and one found a real zero-day inside Hugging Face’s infrastructure. Here is what happened, and why a UK think tank saw it coming.
The Senate Commerce Committee advanced KOSA, the CHATBOT Act, and two more AI chatbot bills for kids on August 5. Here’s what each one requires, and why child safety groups, privacy advocates, and Big Tech all reacted differently.
OpenAI slashed GPT-5.6 Luna prices 80%. Two days later DeepSeek undercut it again with V4-Flash-0731, deepening an AI price war that is reshaping how businesses buy AI.
More News
Apple sent threat notifications to users in 110 countries on August 13, and for the first time the warning lands on the iPhone Lock Screen. Here is what the alert means, how to spot a fake, and the six things to do in the first hour.
Through August 15, GameStop is giving the same trade credit for defective official controllers that it gives for working ones. Stick drift qualifies. Completely dead controllers do not. Here is how the offer actually works.
Researchers at Kyoto University used the orbital drag of roughly 1,200 Starlink satellites to build the first tomographic map of Earth’s thermosphere, turning a commercial broadband network into a planet-scale atmospheric sensor.
An internal Google DeepMind memo tells applicants there is a “non-trivial probability” their CV will be screened out incorrectly, and points them to a separate form that a human will actually read. Google says the system works fine.
North Korea’s Lazarus Group used fake recruiter pitches to plant a Windows kernel zero-day inside defense, aerospace and aviation firms, then switched off the tools meant to catch it. Microsoft patched CVE-2026-68820 on August 11.
OpenAI’s new GPT-5.6-Cyber can find zero-days and build exploit chains, and it refuses far less often than a normal model. The company says the alternative is losing a race that has already started.