Added, by axis
Axes with the most records come first. A record that spans several axes is counted once, under its main axis.
⑩6 records
Safety, Alignment & Interpretability
- Oct 15, 2026Anthropic, METR — Anthropic gave a team from the independent evaluator METR employee-like access — 6 METR employees received company hardware, office space and access to internal tools and systems roughly on par with its own risk assessment teams SourceIt described evaluator partnerships in 3 tiers and presented this as the deepest, saying it is in early discussions with other organizations including Hugging Face. It added the caveat that information about model architectures carries a high leak risk and will be shared only when highly relevant, with internal approval
- Oct 8, 2026Anthropic — Anthropic launched the Anthropic Cyber Mission — starting both the Critical Infrastructure Defense Program, which gives frontier models, on-site engineers and threat research to providers that protect the operational technology (OT) of power grids, water and transport, and OSS Scanner, which gives open-source projects free periodic security scans SourceFounding partners are Accenture, Booz Allen, CrowdStrike, Deloitte, Dragos, Hitachi, Insane Cyber, Nozomi Networks, Palo Alto Networks, PwC and Rockwell Automation. Anthropic said OSS Scanner reports are generated by the model and sent without human review, with an expected true-positive rate above 90%, and forecast that AI will favor defense in 2 years but may not in the near term
- Oct 8, 2026OpenAI — OpenAI reported disrupting 2 AI-enabled 'false front' influence operations — the Russia-origin 'Dark Clark', which co-opted people in Latin America to run a think tank, and the Iran-origin 'Bogus Bylines', which pitched long-form articles under 7 fake journalist personas SourceBoth operations placed some content in mainstream outlets, and OpenAI said Dark Clark is Category 5 on the influence-operation Breakout Scale, the first Category 5 operation it has disrupted since it began reporting. It concluded that AI was mainly used to make old-style operations' workflows (such as drafting internal reports) easier rather than to create new techniques
- Oct 6, 2026Anthropic — Anthropic expanded its Cyber Verification Program (CVP) — folding in Project Glasswing and reorganizing it into 3 access tiers (Defense, Red Team and Specialized), each with different cyber blocking levels and security requirements SourceOn the company's CyScenarioBench test, generally available models were blocked at the first prompt every time, the Defense tier was blocked in 46 of 50 runs, and the Red Team tier succeeded in 34 runs with no blocks (matching the 67.6% success rate without safeguards). Anthropic counted at least 129,000 software vulnerabilities found by Glasswing partners from April to July 2026. The Specialized tier is reviewed together with the US government
⑥5 records
Reasoning Models & the AGI Debate
- Oct 7, 2026Anthropic — Anthropic launched Claude Haiku 5.5 — a small model for high-volume, low-cost work and the company's fastest model, available immediately on all platforms including AWS, Google Cloud and Azure SourceThe company said requests up to 100,000 tokens cost 90% less than Haiku 4.5 and longer requests 50% less, for about 75% lower average running cost. The same day it halved Sonnet 5.5 cache-read prices and introduced a monthly API credit for Max and Team subscribers, encouraging setups that use the small model as a subagent
- Oct 7, 2026OpenAI — OpenAI brought GPT-6 and Intelligent UI to ChatGPT chat — answers combine text, graphics, buttons, forms, charts and interactive elements, rolling out with GPT-6 Sol for Plus, Pro, Business and Enterprise and GPT-6 Luna for Free and Go SourceThe GPT-6 generation reached the default chat of ChatGPT, which has more than 1.2 billion weekly users. Because it begins answering while it keeps thinking, GPT-6 Instant starts answering 44% sooner on average than GPT-5.6 Instant on questions that need web search (company-reported)
- Oct 6, 2026OpenAI — OpenAI released a batch of mathematical results produced by an internal frontier model in a GitHub repository — with revision and citation rules shaped by advice from the IAS advisory group on mathematics and AI, and a $5 million contribution to a new Simons Foundation initiative that helps mathematicians examine AI results SourceIt also published 10 summaries of the model's reasoning, compute estimates expressed as ChatGPT Pro usage and statistics on attempted problems, saying the average result used about 3 hours of ChatGPT Pro thinking. The first case of setting out how AI math results are released to the field after the advisory group was formed in September
- Sep 30, 2026Google DeepMind — Google DeepMind announced Gemini 4 Argon — a frontier model for long-horizon professional work that raises the output token limit from 64,000 to 1 million, rolled out first and in phases only to trusted cyber defenders in the Fairwind Program SourceCompany-reported scores: DeepSWE v1.1 77.9%, AutomationBench 51.3% (first), LVBench 91.7%, CWE-bench v1 68% (tied first). The introductory price is $2/$10 per 1 million tokens (later $4/$20), and Google said it is taking part in the US government's voluntary pre-release access process as it widens access. Internally, Google said Argon agents freed more than 300 TiB through data center memory optimizations
⑤2 records
Loop & Harness Engineering
- Oct 13, 2026Chuyang Wei et al. — Self-Evolving Agent Harness released — formalizes harness self-evolution, which revises prompts, memory, workflows and executable components while model weights stay fixed, and analyzes its reliability with evolutionary search and continual PAC learning SourceUnder explicit assumptions it gives an expected generalization bound, a retention bound conditional on historical regression budgets, and a high-probability rule for admitting revisions using fresh independent data. It concludes that real deployment needs independent safeguards for learning, runtime privileges and consequential actions
- Oct 9, 2026Ruiyao Xu et al. — Harness² released — a framework that combines task-level recursion (Propose-Probe-Compose), which edits the harness for each open-world task without ground-truth verifiers, with harness-level recursion that accumulates that experience into an improvement harness for future tasks SourceThe authors reported that a single step raised GLM-5 and Gemini 3.5 Flash over their base harnesses by up to 11.5 points on WorkBuddy-Bench, 4.5 points on JobBench and 11.8 percentage points on LAB. Research on recursive self-improvement that edits the harness rather than the model continues
⑭2 records
Physical World & Real-World Impact
- Oct 15, 2026OpenAI — OpenAI launched the Rosalind Academic Plan for academic research labs — eligible labs get up to 30 Business Standard seats free for up to 12 months and can use GPT-Rosalind after trusted-access approval SourceIt gives access to all commercially available frontier models within plan limits, continuing the competition among labs to offer free seats to university research labs
- Oct 13, 2026Google — Google made Gemini 4 Argon available to developers — on the paid tier of the Gemini API and in Google AI Studio and Antigravity (for AI Ultra subscribers), and the same day showed how to use Argon across robot development SourceCompany-reported 87% on the embodied reasoning benchmark ERQA. Google said that from a single photo of a real tabletop Argon built a MuJoCo simulation twin on its own and collected more than 4,500 episodes, raising real-robot success from 56% to 75% and cutting completion time from 35 seconds to 24 seconds. A general model taking on robot data labeling, value functions and simulation building
⑨1 record
Open-Weight Lineage
- Oct 15, 2026Meta — Meta released the weights of Muse Spark Open — a natively multimodal model for coding and agents with 816 billion total and 42 billion active parameters and a context of 1 million tokens, on Hugging Face under the Muse Model License SourceMeta claims it is the most capable US open-weights model to date. It adds safety post-training for standalone use on top of the same base model as Muse Spark 1.3, and ships with SGLang and vLLM optimizations and Helion kernels (1.03–3.95 times faster on GB300). Muse Spark, which began as a private preview in April, reached open weights within half a year
⑬1 record
Compute Economics & Geopolitics
- Oct 8, 2026Anthropic — Anthropic committed $150 million over 3 years to the Genesis Mission — a federal initiative led by the US Department of Energy (DOE) to accelerate science with AI, extending Claude to more than 15 participating agencies including NASA, NIH and NSF SourceA follow-up to the DOE and Genesis Mission partnership announced in December 2025. After bringing Claude to national laboratory scientists, it widened the scope to federal science agencies as a public science investment
Showing 14 here; there are 6 more, including 3 other axes. Open the Advanced page to see them all.
How this relates to the main moduleThis page holds
only what was added that week. For the whole story, open
the main module — it is
one latest edition, revised every week, while weekly editions
stay as they were.