Scale AI Rejects European Privacy Complaints Over Freelancer Monitoring
publishers 112articles 365first reported 23 Sep, 14:00 UTCdeveloping since 30 Sep · 33 editions
Scale AI rejected privacy complaints filed against the company in France, Italy, the Netherlands and Austria, calling allegations that it illegally monitors freelance AI-training workers “unfounded and inaccurate.”
Privacy group noyb filed the complaints on Tuesday, alleging that Scale’s Outlier platform unlawfully monitors more than 900,000 freelancers who produce data used to train AI models. The group asked regulators in the four countries to fine the company.
Scale said its platform practices are disclosed to contributors and require their consent. The company said it uses standard tools to confirm identities, prevent fraud, safeguard confidential project information and ensure accurate payments. It also said contributors see pay rates before accepting assignments and that rates vary according to skills, project details and location.
A Scale spokesperson told Reuters that some projects use a separate tool to record time and screen activity. According to the spokesperson, contributors opt into that monitoring and manually activate the tool before beginning work. The spokesperson also said Scale does not sell workers’ personal data to third parties or use that data to train customers’ AI models.
HOW THIS STORY WAS MADE
Written from 8 articles, 5 with the publisher's own text; 8 independent newsrooms once syndicated copies count as one; this version written 666 h after the record first saw the story; the editor kept it as written.
- Publishers
- 112
- Source articles
- 365
- Given to the writer
- 8, 5 with the article's own text
- Independent newsrooms
- 8
- This version written
- 666 h after the record first saw the story
- Second model (editor)
- kept as written
- Publication gate
- passed
- Human review
- none
WHAT THE EDITOR DID
The second model checked the writer's draft against the same sources and published it as written.
Written by a language model from the sources above, then checked by a second model that may only cut, attribute or correct. How it works →