Google's AI finds Android bugs faster than anyone can patch them
Google has told phone makers it will drastically cut Android security backports because its own AI models are discovering vulnerabilities faster than its human teams can fix them.
Cursor's code-execution bug sat unpatched for seven months
A flaw letting any Windows repository run arbitrary code the moment it is opened in the Cursor editor was reported in December, reproduced, acknowledged, and then met with silence across 197 shipped versions.
IBM blames a rival AI model for its earnings miss
IBM missed Q2 expectations and CEO Arvind Krishna named a competitor's frontier model as a cause, saying buyers are pausing cybersecurity deals until they know how much AI changes what security is worth.
Thomson Reuters cuts 500 engineers to hire 250 AI-native ones
Thomson Reuters is eliminating up to 500 engineering roles while planning 250-plus new ones described as senior and AI-native, and its stock rose about 5 percent on the news.
Voters removed a Senate president over a data center
Utah Senate President J. Stuart Adams, the longest-serving in state history, lost his first-ever primary challenge after becoming the face of a fast-tracked data center approval.
What Ring-2.6-1T's model card actually says
Ant Group's openly downloadable trillion-parameter model is real and MIT-licensed, but its benchmark claims are vendor-supplied and measured against a previous generation of rivals -- not the current frontier.
Bonsai puts a 27B model on your phone, and shows what breaks
PrismML shrank a 27-billion-parameter model from 54 gigabytes to 3.9 by storing each weight as a single bit, and published the benchmark table showing tool calling and vision falling off a cliff.
KronQ makes 2-bit models work where the standard method collapses
Adding gradient information to model compression takes a 2-bit LLaMA-3-70B from complete gibberish under the dominant method to a functioning model, suggesting the standard approach has been missing half the picture since 2022.
Hassabis proposes a FINRA for frontier AI
Demis Hassabis published a governance framework calling for a US-led, industry-funded standards body that would review frontier models 30 days before release and could eventually coordinate an industry-wide slowdown.
The zero-cost fallacy: open source under AI pressure
ThoughtWorks argues open source is being ground down by AI-generated pull requests, meaningless popularity metrics, and a licensing model one participant called a profound collective mistake.
Anthropic's doomer ad, and Altman's roast
Anthropic's new ad opens on a burning house and cuts to surveillance footage and rows of tombstones, and Sam Altman spent the day mocking it on X.