coldpress

The State of AI Disclosure 2026: what 1,088 EU sites' chat widgets say

Posted by coldpress on Tuesday August 25, 2026 @04:46PM
from the disclosureproof.com dept.

https://disclosureproof.com/research/state-of-ai-disclosure/

Re:AI Labels Are Big Tech's Most Basic Responsibility, Even Those Clau... on Sunday August 23, 2026 @09:21AM (#49408658)

who benefits from a watermark something like anthropic can only detect?

Re:Show HN: Figranium – Build browser tasks visually, execute via API ... on Sunday August 23, 2026 @09:19AM (#49408641)

does it get through consent walls?

Re:Claude Watermarks AI-Generated Text on Saturday August 22, 2026 @04:14PM (#49403407)

Only Anthropic can check this and a paraphrase strips that. Thats why the AI Act puts the labeling duty on whoever publishes the text not on detection.

Re:Dutch regulator fines Uber €825M for letting AI deactivate driver a... on Saturday August 22, 2026 @04:13PM (#49403393)

Art 22 GDPR doesn't distinguish it says "solely automated processing" Five ifs and a model are the same thing to the regulator.