Interesting.
During this red-teaming exercise, we also found a vulnerability that allows for unauthenticated remote code execution in Brotli, a widely-deployed compression library included in many pieces of software such as Node.js, Bun, and Envoy.
The flaw could let an attacker run their own code on affected systems by sending or replying with malicious compressed data, without any need to create a user account. That is, an unpatched server that accepts compressed input could be taken over by a malicious client, and an unpatched client that accepts compressed input (such as “curl --compressed”) could be taken over by a malicious server.
This root cause issue was actually fixed in the Brotli source code in February, but it wasn’t known at the time that it gave rise to a security vulnerability and there was no more recent release of Brotli. Since Brotli is built into other software, any software which includes this library needs to take the update too, followed by the users of that software. We disclosed the vulnerability to Brotli and other downstream projects such as Node.js, which released or are currently releasing security updates, and Brotli cut new release versions and published a CVE.
More details on the patch and mitigations below:
patchtheplanet.org/brotli
0
0
1
87
RSI OG.
Concrete algorithms for recursive self-improvement (RSI) date back to 1987. 1992-: gradient descent-based neural RSI. 1994-: RSI for reinforcement learning with self-modifying policies. 1997: RSI plus artificial curiosity and intrinsic motivation. 2002-: asymptotically optimal RSI for curriculum learning. 2003-: mathematically optimal RSI through the Gödel Machine. 2020s: new stuff (including RSI survey) people.idsia.ch/~juergen/metal…
0
0
2
154
Great work.
In August, @AISecurityInst reported an incident where Claude Mythos 5 took unauthorized actions on the internet during AISI’s testing of its cyber capabilities.
We've now produced a further analysis of this incident, to better understand the model's alignment properties.
anthropic.com/research/align…
0
0
0
57
Agentic AI coding.
Git was built for developers. What does it look like when AI agents start writing the code?
@ashtom, former GitHub CEO and now co-founder and CEO of @EntireHQ, is taking the AGNTCon + MCPCon NA stage with “Building the Git Forge of the Future.”
Live now: bit.ly/4dVkjdO
0
0
1
126
Very cool - TorchTPU is now open source: Run native PyTorch on TPUs. cloud.google.com/blog/topics/ai…
0
0
0
47
Very nice.
Google Cloud and @AnthropicAI now offer two ways to build with Claude models.
Choose Anthropic managed to access the full Claude Platform with Google Cloud IAM and billing; select Google managed for in-boundary security, regional data residency, and native cloud integrations.
Y
0
0
0
44
Teams of agents were assembled, powered by OpenAI's most cyber-capable models, and tasked with finding weaknesses in the OpenAI research environments.
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing…
0
0
0
71
Valuable information from when a METR team were embedded in OpenAI from August this year. Here is their report.
Read our public overview for more findings, limitations, and design ideals:
metr.org/blog/2026-10-2…
0
0
0
77
This is great news!
OpenAI models are coming to Google Cloud.
We're bringing our frontier models and Codex agentic workflows to Google Cloud's Model Garden, so enterprises can use our technology where they already build.
openai.com/index/openai-m…
1
0
0
84
Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API. Open weights by Nov 4, 2026.
Kimi K3.1 Pro launches at 4th on the Artificial Analysis Intelligence Index scoring 55, making large gains in agentic performance; it will be the leading open-weights model once its weights are published
@Kimi_Moonshot's new reasoning model gains 11 points over Kimi K3 (max, ude Opus 5.5 (max with fallback, 58), Claude Sonnet 5.5 (max with fallback, 56), and Muse Wonder Full (max, 56) on the Intelligence Index.
Once the weights for K3.1 Pro are released it will be the highest-scoring open weights model and the gap between the leading open weights and proprietary models on the Intelligence Index will narrow from 12 points to 3.
Key results:
➤ Large gains in agentic performance: Kimi K3.1 Pro scores 3rd on GDPval-AA with 1,846 Elo, up 312 Elo from Kimi K3, behind only Claude Opus 5.5 (1,866 Elo) and Muse Wonder Full (1,857 Elo). Kimi K3.1 Pro also achieves the highest score of any model on Terminal-Bench 4.0. K3.1 Pro scores 67%, up from 13% for Kimi K3 and ahead of Claude Sonnet 5.5 (64%). AutomationBench-AA, which tests agentic SaaS workflows, rises from 58% to 73%.
➤ Gains in other evals are more moderate: Humanity's Last Exam rises from 47% to 52%, CritPt from 23% to 25%, and SciCode from 59% to 61%. AA-Omniscience Index rises from 20 to 30, as the hallucination rate falls from 53% to 44% and accuracy rises from 48% to 51%.
➤ Cost per Task roughly equal to Kimi K3 at $1.98 per Task: K3.1 Pro uses 55K output tokens per Intelligence Index task, 14% more than Kimi K3 at 48K, while its first-party token prices are 33% lower.
Additional model details:
➤ Context window: 1M tokens.
➤ Size: 2.8T total, 104B active parameters
➤ Input modalities: Text and image
➤ Weights and license: Weights currently not published, Moonshot is expected to release the weights soon after launch
➤ Pricing: $2/$10/$0.20 per 1M input/output/cache hit tokens, 33% lower than Kimi K3 on all three rates.
Congratulations to the team at @Kimi_Moonshot on the release!
0
0
0
74
Exclusive: OpenAI has acquired Orchestra Bio axios.com/2026/10/20/exc…
0
0
1
66
Peter Morgan retweeted
Major U.S. financial institutions are facing “unvetted and unauthorized AI agent activity,” says @MadisonMills22, a senior AI reporter at Axios, calling it “a nightmare” for banks.
Mills and CBS News’ @TheKellyOGrady share what exactly is happening with the agents — and how to unts.
Mills and CBS News’ @TheKellyOGrady share what exactly is happening with the agents — and how to unts.
3
11
23
9,786
Wow, nice!
New on our Frontier Red Team blog: using Claude to formally verify software at large scale.
Formal verification can prove that code is free of certain bugs, but historically it has only been feasible for small codebases (or required a lot of expert time).
With AI models, however, this is beginning to change. In collaboration with @theoremlabs, we used Claude to prove security properties in over one hundred real-world open-source libraries over the span of a few weeks.
Read our research in full here: anthropic.com/research/forma…
0
0
0
53
Context is king, when it comes to building reliable enterprise agents. neo4j.com/blog/news/neo4…
0
0
0
68
Nice.
Today we’re announcing a new collaboration with @nvidia to bring AI to industrial engineering:
- faster simulation
- open benchmarks and supporting datasets for physics AI
- evaluations for agentic AI in CAE and EDA
More here: mistral.ai/news/mistral-n…
0
0
0
83
Pareto frontier.
MiniMax M3.1 Flash: Space Performance + Bunny Speed
🚀🐰 Yes, Space Bunny - which has topped global leaderboards with 200T+ tokens - was an earlier version of M3.1 Flash.
💰Redefining the performance economics of agentic AI. M3.1 Flash achieves frontier-level intelligence s Cowork (Automation-Bench: 62%; ALE: 30%), Coding (DeepSWE: 74%) and Visual Reasoning (BabyVision: 91%), at 1/3 the cost of M3 and 3x the speed.
API pricing (flat rate): $0.15 / $0.60 per 1M input/output tokens, with cache reads at $0.015 / 1M tokens.
Model weights will arrive on Oct 23.
API: platform.minimax.io
M Plan: minimax.io/m-plan
MiniMax Code for desktop: agent.minimax.io/download
0
0
0
73
TorchNeuron supports Trainium2 and Trainium3. It will be released as open-source under the Apache 2.0 license on October 20 #pytorch #trainium #aicompiler.
aboutamazon.com/news/innovatio…
aboutamazon.com/news/innovatio…
0
0
0
85
Harness Self-Evolution - Theory, Algorithms, Evaluation, Applications, and Impact.
zgcagi.github.io/Harness-RSI/
zgcagi.github.io/Harness-RSI/
0
0
1
88