AINewsnow

Epoch AI test finds no frontier model meets its employee standard

This story is from 2026-11-17. It is preserved in the archive; the latest stories are on the live feed.

Epoch AI tested Fable 5.2, GPT-6.2 Astra, and four other frontier models on real workplace tasks; none met the benchmark's employee bar, including the two top scorers on its ECI.

Read the full story at Epoch AI ↗

Timeline · 2 reports

  1. 2026-11-17 21:44 · AlphaSignal
    Epoch's Benchmark Catches Fable 5.2 and GPT-6.2 Astra Failing Real Work
  2. 2026-11-17 00:00 · Epoch AI
    How good are Fable 5.2 and GPT-6.2 Astra at real work?

More stories

  1. Microsoft CEO warns AI-driven bug discovery is raising the pace of tech outages — OpenAI News
  2. ChatGPT, Claude, Gemini and Grok helped plan terror attacks in US government test — The Independent Tech
  3. OpenAI unveils Performance Factory, a Codex-powered system that keeps ChatGPT fast — OpenAI News
  4. ChatGPT and Gemini tested on holiday gift recommendations for a drummer who loves vinyl — Tom's Guide AI
  5. OpenAI launches accounting plugin for ChatGPT to streamline business financial management — Economic Times Tech
  6. OpenAI launches ChatGPT for Sales and Accounting with new partner plugins — OpenAI News
  7. OpenAI Expands ChatGPT for Healthcare With 20+ Plugins and Computer-Use Agents — OpenAI (X)
  8. UK government testers found OpenAI's pre-release GPT-6.2 Astra easier to jailbreak than predecessor — VentureBeat AI

Read the Daily Brief: the day's most important AI stories →