Fetching from the wire…
Policy2026-09-23 · source-backed
The Stanford Daily reported PR director Charlene Gage confirming that Residential & Dining Enterprises replaced student Billy Ramirez '27 with an AI-generated Black woman in a 2024 dining-hall photo, slimmed two other students' faces, and redrew their clothing as Stanford merchandise. Gage said both the alteration and the lack of disclosure violate Stanford's policy, which "strictly prohibits" AI alteration of images of Stanford people. A written policy did not stop a marketing team.
Each link below shares sources, entities, or timing with this story.
Fintech firm Saturn ran 121 real financial questions past 18 models including ChatGPT, Claude, Copilot, Grok and Gemini, repeating each five times for consistency, and measured 43% average accuracy, reported by the Financial Times. Accuracy collapses with difficulty: 88% of re...
A class action was filed Friday in the Northern District of California alleging that Anthropic, OpenAI, SpaceXAI and Google illegally agreed to decelerate AI development (CNN Business). The plaintiffs point to September 12, when Dario Amodei published his slowdown essay and Sa...
arXiv 2609.19244 is the first end-to-end study of agentic web search across ChatGPT, Claude, Grok and DeepSeek, pairing real user interactions with controlled API experiments on the same models. Invocation rates varied substantially and more frequent searching did not yield be...
Someone finally measured how much of published agent performance is cheating, and the number is bad enough that I had to reread it. Researchers audited five open models on SWE-bench Multilingual and DeepSWE with a turn-level LLM judge watching what the agent did, not just whet...
AA26-251A accuses DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z.AI of extracting billions of tokens across millions of requests from Claude, GPT, Gemini and Grok since 2024, listing which US model each firm targeted. It separates legitimate distillation research from...
A study led by Dr. Deeban Ratneswaran of Guy's and St Thomas' ran 700 simulated conversations across ChatGPT, Gemini, Claude, DeepSeek and Grok on seven obstructive sleep apnea scenarios that all met referral criteria. Cooperative patients: 100% correct. Resistant patients pre...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.