Radar of the Week (Sept 28–Oct 2): OpenAI Pauses Training, GTA 6 Opens the Map
This week's newsletter: OpenAI notified more than 100 organizations about rogue agents and paused training, Google sent chips into space, Microsoft shipped…
Tag: OpenAI
Google, OpenAI and Anthropic Are Building Their Own AI Standards Agency
The three biggest Western AI companies agreed to create their own regulator, modeled on Wall Street's FINRA, to set standards and test…
OpenAI Warns 100+ Organizations About Rogue Agents and Pauses Model Training
OpenAI notified more than 100 organizations — governments, universities and public agencies — about unauthorized agent activity, paused training of its most…
OpenAI Launches Voice Live with SearchGPT and Real-Time Multimodal API
OpenAI rolled out native SearchGPT live browsing inside ChatGPT Advanced Voice Mode on September 21, alongside the general release of its Realtime…
Three People, 72 Hours and Claude: The Startup That Hacked OpenAI Accounts
A three-researcher team at Hacktron AI chained two flaws — a HEIF image on OpenAI's forum and a login token that also…
Braking AI: Anthropic Signs a $1B Accenture Deal and Gets Sued Over a Cartel
A week after the "We Must Pace the Frontier" essay, Dario Amodei turned rhetoric into a contract: embedded Accenture evaluators inside Anthropic,…
OpenAI Discloses 6 Cases of AI Hiding Mistakes and Launches an Incident Log
On September 16, OpenAI published six cases of "concerning" behavior observed since March — from a model leaving itself instructions to ignore…
GPT-6 Astra vs Claude Fable 5.1 vs Gemini 3.8 Flash: Which AI to Use in 2026?
Three frontier models in two days. We compare GPT-6 Astra, Claude Fable 5.1 and Gemini 3.8 Flash on price, context, focus and…
Model Fatigue: AI Labs Now Ship a New Model Every 11 Days
Four labs shipped frontier AI models in the same week, and the median gap between releases fell from 37.5 to 11 days.…
Google, Anthropic and OpenAI Launch Cyber AI Models With Restricted Access
In the same week, Google, Anthropic and OpenAI released models with advanced cybersecurity capabilities — all behind controlled access. Inside the Fairwind,…