ENTIA Press Room · status as of 12 September 2026

A machine-readable rights reservation, an invitation from the European Commission, and the silence of the large labs

ENTIA maintains 11,330,391 entities verified against official registers across ten European countries, every field anchored to its source and date. Since 19 July 2026 that corpus carries a text and data mining rights reservation on every response. Between that date and 1 September, 170 crawler families made 2,279,997 requests, every day sealed, the record signed. The European Commission invited ENTIA to contribute in writing. Twelve operators were notified with their file and a licence proposal. Six have replied. None of the five GPAI Code signatories has answered the question on machine-readable reservations: Anthropic and Microsoft replied without addressing it; OpenAI, Google and Amazon have not replied.

Read the case Status by operator
Live

The Machine Crawl Census, live

Which AI operators, search engines and crawlers are reading ENTIA's corpus today: logos, daily charts and the 170 families of the signed record. Updated every day.

Open the live census
2,279,997third-party machine requests19 Jul to 1 Sep 2026 · consolidated record signed on 2 Sep
12operators notified31 Aug and 1 Sep · sealed file and licence proposal
6 / 6have replied / have not replied2 remediated · 1 opened a channel · 3 did not address the question
0 of 5GPAI Code signatories that answered the questionAnthropic, OpenAI, Google, Microsoft and Amazon
Live case

What the case documents

Updated 12 September 2026

What is happening

ENTIA has not asked anyone to stop crawling. It expressed a machine-readable rights reservation, measured access at its own edge, sealed every day, signed the record with a qualified electronic signature and notified each operator through the channel that operator itself publishes, with a licence proposal. On 24 August the European Commission, DG CNECT, copyright unit, invited ENTIA to submit a written contribution; it is published at entia.systems/eu-submission. Two SEO tools remediated. One opened a commercial channel. Anthropic replied pointing to robots.txt. OpenAI, Google, Meta, Apple and Amazon have not written.

Methodological boundary

The materials document access, delivery, volume, frequency and persistence, attributed by agent name and originating network together. They do not assert training, RAG, grounding, storage, infringement or any other downstream use that has not been directly observed. Nor do they assert that the European Commission has assessed or endorsed anything.

The case

The reservation, the Commission, the signed figures, the silence of the labs and the full timeline.

Status by operator

Twelve files with dates, names, channel used and what each one said.

Contribution to the European Commission

The written contribution of 24 August, the update of 2 September and the signed record.

Press contact

Interview requests, verification of figures and access to the files.

Five-Layer Framework

Research methodology: access, eligibility, retrieval, citation and influence. Version 0.2 (16 September 2026), CC BY 4.0.

DOI 10.5281/zenodo.22795758

Press kit and assets

Complete downloadable press kit, biography, logos and brand kit.

Archive: press pack of 31 July 2026

The first press pack (European Machine Access Case). It describes the 7 to 27 July window and is superseded by the current case.