Bounty hits $3,000 as researcher shows AI text detectors are trivially evadable
ctjlewis · x · 2026-11-16
Responding to challenges over whether his bounty targets false positives, ctjlewis clarifies: the reward for proving an AI-text detector false positive (human writing flagged as AI) has been bumped from $500 to $3,000, but the target must be a detector like Max or Pangram submitted by someone he trusts.
He notes he has so far only demonstrated false negatives — AI text slipping past detection — to show that evasion is trivial, and says finding a false positive is still possible, just more expensive. The exchange underscores how fragile current AI text detection remains under adversarial pressure.
Related event: Bounty for Beating AI Text Detectors Raised to $3,000(2 posts)→
More from Safety
- BIS Issues Interim Final Rule Addressing Remote Access to Advanced AI Chips — apples_jimmy · 2026-11-16
- Fed Governor Michael Barr: Banks Are Already Catching AI Agents Taking Unauthorized Actions in Customer Accounts — Polymarket · 2026-11-16
- US AI export control carves out commercial recommender systems, sparing ByteDance's search algo — zephyr_z9 · 2026-11-16
- Eval Firm Trajectory Told 40 Agents "NEVER Stop" in Harness Without Telling Epoch — peterwildeford · 2026-11-16
- AI-text detection claims need mechanistic interpretability reports, not post-hoc explanations — madeofAjala · 2026-11-16
- No legal liability framework exists for model misalignment harms, and that's a disaster — LuizaJarovsky · 2026-11-16