Friday, October 30, 2026
A malicious update to Anthropic’s VS Code extension was distributed through Cursor’s extension marketplace for several hours, potentially reaching up to 5,000 developers. The update tried to give attackers remote access to their computers; neither company has explained how it reached Anthropic’s official account.
Free daily newsletter
Get these stories in your inbox every morning.
One email a day. No spam. Unsubscribe anytime.
Subscriptions open soon. Check back shortly.
US tech stocks fell sharply for a second straight day as investors questioned whether heavy AI spending would pay off. Intel, Alphabet, Amazon and Reddit were among the companies whose shares declined, while the Nasdaq dropped more than 3.5%.
OpenAI has revised its public guidelines for GPT models, emphasizing honesty, safety and respect for users’ wishes. The rules tell AI agents to explain their assumptions and actions and accept human oversight, though past incidents show that models do not always follow safety measures.
In tests by Epoch AI, the open-weight model GLM 5.3 built Firefox attacks without human help in less than 48 hours. It used clues from fixes not yet delivered to users to create a webpage that could steal local files; Mozilla has since fixed a flaw the attacks relied on.
Google’s Suncatcher satellite is operating in orbit and has used an onboard AI model to write a poem. Google is testing how its AI chips handle space conditions before deciding whether to launch more satellites.
You do not have to believe in a god to seek a religious exemption from using AI at work. The University of Kentucky granted administrator David Durman’s request to avoid AI because he says it conflicts with his humanist beliefs.
The Wikimedia Foundation filed a brief in The New York Times’ copyright case against OpenAI and Microsoft, asking the court to take a careful approach to fair use. It supports some AI training and use of outside sources, but warns against systems that reproduce articles without credit.
Cisco has given engineer Padmanabhan Krishnamurthy unlimited AI tokens to help rewrite 98% of its older code within two years. He has four weeks to show that his approach can work.
Foxconn aims to raise its share of worldwide AI server sales from about 40% to 50% within a couple of years. The company also plans to expand U.S. manufacturing as demand for AI equipment grows.
Anthropic could receive a high junk credit rating after its planned IPO, according to a Financial Times report on investor Blackstone’s forecast. Blackstone had previously expected the AI company to receive an investment-grade rating before 2027.
Free daily newsletter
Join readers who get Training Data Daily before their first meeting.
One email a day. No spam. Unsubscribe anytime.
Subscriptions open soon. Check back shortly.