Topic
Vendor obligations arising from EU rules on marking AI-generated content, including Code of Practice commitments and transition periods.
Techdirt argues Anthropic complied with the EU AI Act more broadly than the law required. For anyone shipping model-written text, the vendor's implementation is the constraint that actually binds.
Reality
EU rules require AI firms to mark generated text by 2 December, and vendors will comply. Researchers say light edits strip the marks, so any policy built on detecting them is built on sand.
Reality
Anthropic began marking Claude's output on 2 August. Within days an open-source stripper had more than 14,000 GitHub stars, which settles what the mark can and cannot prove.
Perspective Coverage
3 publishersReality
Anthropic will label Claude's text and files to satisfy the EU AI Act. That fixes provenance for one vendor's output and leaves the hard part of assessment policy where it was.
Reality
The AI Act's tagging duty is live. The rules that would make biometric verification vendors prove they resist manipulation are not, and fraud does not label itself.
Reality
Anthropic described statistical text watermarking for Claude on August 14 but has published no detector, so neither the mark's durability nor its removal can be independently checked.
Reality
Google's budget tier now handles the summarize-and-compact work that fills agent invoices. The 75-cent introductory input rate lapses on December 31, 2026, and then input goes back to $1.50.
Reality
Article 50 and California's transparency act both switched on 2 August 2026. The watermarking work belongs to model vendors; a shipping team owes chatbot, deepfake and public-interest text notices.
Reality
The mark lives inside token sampling, is invisible to readers, and needs a key to read. Provenance testing just became an operational question for DLP and insider-risk teams.
Reality