Anthropic's watermarking and the shift to AI governance and authority
asusarla · x · 2026-08-18
The post discusses Anthropic's implementation of watermarking to comply with the EU AI Act (Article 50) and its limitations regarding accountability.
- Watermark Limits: Watermarks prove a model touched the content but fail to establish who authorized the action, the scope of permission, or ownership of the decision. Liability shifts to the deployer.
- Governance vs. Provenance: Provenance shows machine involvement; governance demonstrates authorized intent.
- Solution: AptlyDone proposes governed identities for agents, featuring defined scopes, enforced limits, expiration, named owners, and proof of authorization.
The discussion highlights that technical provenance (watermarks) is insufficient without governance-based authority proof.
More from Models
- Prediction: frontier AI revenue growth could slow even as AI usage keeps rising — yi111 · 2026-10-16
- Meta teased 816B-param open-weight Muse Spark Open with 1M context, says meme thread — alexandr_wang · 2026-10-16
- Nolla Health Post-Trains Nemotron 3.5 VL: Visual Diagnosis Jumps from 35% to 66.8% — LoganGrasby · 2026-10-16
- Dev argues coding models now beat nearly all engineers in every branch of software — zack_overflow · 2026-10-16
- Prediction: Chinese frontier models could match GPT-6 by March 2027 — teortaxesTex · 2026-10-16
- Liquid AI's d1-ultra, #1 on Artificial Analysis Decision Index, lands on Vercel — JosephJacks_ · 2026-10-16