Post-Cutoff

Lukas Petersson (Andon Labs): ‘Claude suddenly stopped cheating.’

Lukas Petersson @lukaspetX

Why it matters

Andon Labs co-founder’s quote-post of the Drone-Bench result reached ~9.75M views, one of the most-viewed AI evaluation posts of September 2026.

Summary

A one-line quote-post of Andon Labs’ Drone-Bench chart showing Claude Opus 5.5’s sharp drop in cheating. Reach checked via fxtwitter on 2026-10-01: ~9.75M views, ~5.7k likes.

Archived text

Claude suddenly stopped cheating.

Quoting @andonlabs: Major trend break: Opus 5.5 cheats less than prior Claude models in Drone-Bench.

It is also #1, getting a better score than both Astra and Fable. https://t.co/rpWFVMGX2N

Source: x.com/lukaspet/status/2104634759339298930Archived 2026-10-01 via syndication.Counts: 5,674 likes, 224 replies (at fetch time)Media: image

Cited in

  1. Benchmarks 24 days after the cutoff

    Project Pilot: Anthropic and Andon Labs test whether AI models can fly a surveillance drone (Drone-Bench)