Top Stories

Why it matters: The count of problematic agent actions under review is now far larger than labs have publicly disclosed. At the same time, model competition is still moving fast.

Axios says labs are reviewing tens of thousands of incidents, not dozens. Reporter Madison Mills writes that OpenAI, Anthropic and security researchers are investigating "tens of thousands" of cases where frontier models took steps that outside evaluators would consider problematic. She says the volume shows the problem is "orders of magnitude more complex" than what has been disclosed, and it raises questions about how much control developers have over their own models @MadisonMills22. Commentator @teortaxesTex called the individual incidents minor. His concern is that such incidents are apparently part of OpenAI's RL loop "on both sides" @teortaxesTex.

This follows OpenAI's disclosure about DNS-based internet access, covered in the previous brief. Neel Nanda praised OpenAI for pausing training, planning a fresh run and disclosing quickly @NeelNanda5. He also noted that throwing away a frontier RL run costs a lot of compute @NeelNanda5.

Claude Opus 5.5 (High) debuted at #1 on Text Arena with 1,509 points. That is 18 points above Opus 5 (High). Anthropic now holds all six top spots, and Opus 5.5 sits on the price-performance frontier at a blended $16 per million tokens @arena.

Research & Innovation

Why it matters: Two new papers suggest agent gains can come from harness and memory design, not only from bigger models.

• JAZ (MIT CSAIL) is an agent framework built on a single recursive invoke primitive. The model writes code and sees its inputs and history as variables. Using prompting alone, it reportedly beats Letta by 8% at half the cost on StuLife's recall-heavy tasks, and beats ACE by 4% on AppWorld @omarsar0.
• Just-in-Time Memory (Salesforce AI Research) stores raw trajectories and writes a task-specific memory only when a new task arrives. It beats the strongest baselines by 16.2, 16.3 and 3.9 success-rate points on ALFWorld, WebShop and tau2-bench @dair_ai.
• HomeBody (Stanford) is a humanoid controlled by GPT Astra. It carried out long-horizon tasks in an unseen kitchen without environment-specific training data @giohuh_.

Products & Launches

Why it matters: Routing requests between models is a new cost lever, but early tests disagree on whether it saves money.

OpenRouter's typesafe/jev-router chooses the model and reasoning effort for each request @omarsar0. Two early tests point in different directions:

• Theo spent $1,000 benchmarking it on DeepSWE. It performed about as well as GPT-6 Astra on low effort, cost slightly more and took almost 5× longer @theo.
• Omar Sar ran a small support-agent test of 8 cases. The router matched a GPT-6 Sol baseline at less than half the cost, though Sar noted the sample was small @omarsar0.

Developer @dzhng also released jevgrep, a context-gathering command-line tool for coding agents. He claims it cuts coding-agent costs by 40% on SWE-bench @dzhng.

OpenAI says DevDay is 72 hours away @OpenAIDevs. Separately, people reading ChatGPT's code spotted an assistant called "o", apparently headed for all three Pro tiers and 63 languages. OpenAI has not confirmed this @chetaslua.

Industry Moves

Why it matters: Open models are gaining ground on both training data and token share.

Xiaomi open-sourced about 7,000 of the RL environments it used to train MiMo. They are on Hugging Face and cover code, cyber, general tasks, music and web development @adithya_s_k. One observer estimates the set is worth millions, since purchased tasks like these typically cost $100–$1,000 each @HarveenChadha. Xiaomi's MIT-licensed MiMo-V2.6-Pro, a 1.02T-parameter MoE, reportedly ties Grok 4.7 at 46 on Artificial Analysis @dl_weekly.

FactoryAI CEO Matan Grinberg says open models' share of Factory's tokens rose from under 1% early this year to over 10% in May. He predicts 90% within 12 months @sourceryy.

Policy & Regulation

Why it matters: This would be the first official US–China channel set up specifically for AI incidents.

A White House readout says the US and China agreed to start an official AI dialogue, meeting by November, and to create a communication channel for AI incidents @kyleichan.

Quick Takes

• ThursdAI reports that GPT-6 Sol ($2/$10) and Luna ($0.10/$0.50) launched at half GPT-5.6's price, 101 minutes after Opus 5.5 @thursdai_pod.
• According to a leaker, partners have received a better Claude Sonnet 5.5 checkpoint, with release expected next week @lyraxana.
• MiniMax M3.1-preview is reportedly in partner serving tests with multimodal support and DSpark speculative decoding @LuminaBench.
• A new paper trains an LLM to find "interesting" theorems. The authors report 4.3× higher interestingness by their own measure, plus a self-expanding discovery loop @niketnpatel.