| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | AMIE matches doctor diagnoses in 90% of 100 patient interactions with 0 safety stops ↗ source | 2026-10-08 |
|---|
| L1Assisted | L1 | T3Vendor self-report | Astrophysicist uses Claude Science to build first complete ultraviolet sky map ↗ source | 2026-10-08 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Ignite Vision builds AI creator matching on Qwen, cutting screening from hours to minutes ↗ source | 2026-10-08 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Claude Haiku 5.5 generally available in GitHub Copilot ↗ source | 2026-10-07 |
|---|
| L1Assisted | L1 | T3Vendor self-report | Google Research shares RCT findings on AI and professional expertise ↗ source | 2026-10-07 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Gemini in Google Docs can convert briefs into presentation decks ↗ source | 2026-10-07 |
|---|
| L1Assisted | L1 | T3Vendor self-report | GitHub releases ReviewBench for AI code review agents ↗ source | 2026-10-06 |
|---|
| L1Assisted | L1 | T3Vendor self-report | Meta shares six papers from mathematicians working with Muse Spark ↗ source | 2026-10-02 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | PCG uses Seedance 2.5 for Saudi Arabia Railways film ↗ source | 2026-10-02 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Meta releases Immersive Web SDK version 1.0 ↗ source | 2026-10-01 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Google science AI model ranks highest in CDC flu forecasting ↗ source | 2026-09-30 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Anthropic Science Blog: Claude can do Nine Loops ↗ source | 2026-09-25 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Remento uses Google Workspace with Gemini for customer insights ↗ source | 2026-09-25 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Harvey uses GPT-6 Astra for structured legal drafts ↗ source | 2026-09-24 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Stitch supports brand kit and product shots in prompts ↗ source | 2026-09-24 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Gemini in Sheets builds spreadsheets from prompts ↗ source | 2026-09-21 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Alibaba releases Qwen3.8-LiveTranslate real-time interpretation model ↗ source | 2026-09-19 |
|---|
| L1Assisted | L1 | T3Vendor self-report | Google releases Android Bench 2.0 for long-horizon engineering tasks ↗ source | 2026-09-18 |
|---|
| L1Assisted | L1 | T3Vendor self-report | ChatGPT now available in Microsoft Word ↗ source | 2026-09-17 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Cognition's Devin uses GPT-6 Astra for test generation ↗ source | 2026-09-15 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Fortal Studio uses BytePlus for generative short-drama production ↗ source | 2026-09-15 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | MiniMax H3 demonstrates pencil drawing capability ↗ source | 2026-09-15 |
|---|
| L2Partial automation | L2 | T2Independently corroborated | Venice AI uses Wan 3.0 in real creative workflow ↗ source | 2026-09-15 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | MiniMax H3 achieves over 2x real-time denoising on 8x B200 ↗ source | 2026-09-14 |
|---|
| L1Assisted | L1 | T3Vendor self-report | GitHub legal operations manager builds internal tool with Copilot CLI ↗ source | 2026-09-14 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Perplexity engineer uses GPT-6 Astra in Codex for end-to-end testing ↗ source | 2026-09-14 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Yelp uses GPT-Live-1 for natural restaurant reservation calls ↗ source | 2026-09-11 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Gemini in Google Workspace can extract data from Drive folders into Sheets ↗ source | 2026-09-11 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Google DeepMind uses AI to reconstruct memory for documentary 'Love, Rendered' ↗ source | 2026-09-11 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Hailuo AI game trailer teaser ↗ source | 2026-09-11 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Meta highlights Muse Code and Immersive Web SDK for VR development ↗ source | 2026-09-10 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | OpenAI publishes GPT-Live-1 production voice agent benchmark results ↗ source | 2026-09-10 |
|---|
| L1Assisted | L1 | T3Vendor self-report | OpenAI shares cyber defense architecture and playbook ↗ source | 2026-09-09 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Grok adds Coinbase connector for trading and portfolio management ↗ source | 2026-09-09 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | GitHub Copilot can approve pull requests after review ↗ source | 2026-09-09 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Muse Spark 1.3 now available in Cursor ↗ source | 2026-09-08 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | FOX Sports used AI for World Cup broadcast coverage ↗ source | 2026-09-06 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Virgin Media O2 deploys multi-agent system on Google Cloud ↗ source | 2026-09-05 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Claude completes first formalized proof of Fermat's Last Theorem ↗ source | 2026-09-04 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Microsoft reports HydraFusion results on Terminal-Bench 2.1 ↗ source | 2026-09-04 |
|---|
| L2Partial automation | L2 | T2Independently corroborated | Cognizant uses Google Workspace and Gemini Enterprise to cut dev timelines ↗ source | 2026-09-04 |
|---|
| L1Assisted | L1 | T3Vendor self-report | Wan and GMI Cloud AI filmmaking challenge in Taipei ↗ source | 2026-09-04 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | OpenAI introduces GPT-6 Astra ↗ source | 2026-09-03 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Gemini 3.8 Flash available in GitHub Copilot ↗ source | 2026-09-03 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Qwen introduces E-Commerce Bench benchmark ↗ source | 2026-09-03 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Qwen3.8-Max-0902 ranks #1 overall on Code Arena ↗ source | 2026-09-02 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Qwen3.8-Max-0902 tops CodeArena WebDev at 1691 ↗ source | 2026-09-02 |
|---|
| L2Partial automation | L2 | T2Independently corroborated | LatchBio evaluates Grok 4.6 on biosecurity tasks ↗ source | 2026-09-01 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Qwen3.8-Max tops CommerceAgentBench among open-weight models ↗ source | 2026-09-01 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Wan3.0 ranks #1 on Artificial Analysis video leaderboards ↗ source | 2026-08-27 |
|---|
| L1Assisted | L1 | T3Vendor self-report | Developer ships npm library built with Codex ↗ source | 2026-08-26 |
|---|
| L3Conditional automation | L3 | T2Independently corroborated | Turing uses Gemini to automate 60k+ HR tickets ↗ source | 2026-08-26 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Qwen3.8-27B ranks #1 among open models on Image-to-WebDev Arena ↗ source | 2026-08-26 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | MiniMax-M3 cited as cheapest model on agent benchmark ↗ source | 2026-08-25 |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Qwen3.8-27B ranks 9th overall on Code Arena ↗ source | 2026-08-25 |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Cooley builds confidential redaction agent on Gemini Enterprise for Legal ↗ source | —no update mentioned it |
|---|
| L2Partial automation | L2 | T3Vendor self-report | Hitachi deploys agents to frontline workers via HMAX for 200+ use cases ↗ source | —no update mentioned it |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | On migrates 24 core enterprise services to Google Cloud with agent-led strategy ↗ source | —no update mentioned it |
|---|
| L2Partial automation | L3↓ T3 allows only L2 | T3Vendor self-report | Gemini ML skills write and test PySpark code in notebooks ↗ source | —no update mentioned it |
|---|