Copyright can restrict use of the scraped works: Considering | AIGP
7-day money-back guarantee — full refund within 7 days of purchase if you've completed under 20% of the questions. See pricing →
Certifications Tools Flashcards Career Paths Exam Guides Blog Pricing For Teams About

Language

✓ EnglishDeutschEspañolFrançaisPortuguês
Check readiness — free →

Copyright can restrict use of the scraped works: Considering intellectual property law and the EU AI Act

AIGP Understanding How Laws, Standards and Frameworks Apply to AI Hard

Copyright and licence terms can both limit training use, and the EU AI Act obliges GPAI providers to keep a copyright policy and publish a training-data summary.

The question

A startup scrapes copyrighted articles and licensed image libraries to train a general-purpose model it will offer in the EU, and its terms let customers generate derivative images. Considering intellectual property law and the EU AI Act together, which statement most accurately frames its exposure?

Preparing for AIGP? Take the free 5-min readiness quiz →

  1. Once articles are publicly accessible online, copyright no longer restricts their use for model training, though the image licences may still cap how many derivative outputs each customer can generate.
    Almost right on licences, but public accessibility does not extinguish copyright, so scraping protected works for training can still infringe absent a lawful basis or exception.
  2. Copyright can restrict use of the scraped works for training, licence terms may separately limit permitted uses, and as a GPAI provider it must adopt a copyright policy and publish a training-data summary.
    Correct: IP law can limit training use, contractual licences add their own constraints, and Art. 53 requires GPAI providers to have a copyright policy and publish a sufficiently detailed training-data summary.
  3. Because the model only outputs new images, no training input can infringe copyright, and the EU AI Act imposes copyright duties solely on deployers who commercialize the generated derivative works.
    Plausible but infringement can arise from the training use of inputs, and the Act places the copyright-policy and training-data-summary duties on GPAI providers, not only deployers.
  4. Trademark rather than copyright is the governing regime for the training data, so a filed trademark clearance plus a general fair-use notice fully resolves both the initial scraping and the derivative-output risks.
    Plausible-sounding but training-data disputes turn on copyright and licensing, not trademark, and a fair-use notice does not by itself resolve infringement or the Act's GPAI duties.
The trap
Assuming that publicly accessible content is free to use for AI training.

How to remember it

Copyright and licence terms can both limit training use, and the EU AI Act obliges GPAI providers to keep a copyright policy and publish a training-data summary.

How many of these would you get right?

One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.

Test your AIGP readiness — free

More Understanding How Laws, Standards and Frameworks Apply to AI questions

Part of the Certsqill AIGP question bank · Understanding How Laws, Standards and Frameworks Apply to AI · Every answer, right and wrong, comes with its own explanation.