Thousands of people work remotely on AI training platforms, reviewing model answers, writing prompts, and completing specialist tasks. But how much information do these platforms collect about the people doing that work?
That question is now part of a European privacy dispute involving Outlier, the contractor platform operated by Scale AI.
On October 27, 2026, digital-rights organization noyb announced five complaints against Scale AI in France, Italy, the Netherlands and Austria. The group alleges that Outlier subjects contractors to excessive online tracking, intrusive identity checks and inadequate responses to personal-data requests.
These are allegations under investigation, not findings that Scale AI broke the law. Scale AI has disputed the claims.
Five complaints, four countries
The complaints were filed in collaboration with the organizations Intérêt à Agir, DiPLab and Reversing. Works. According to Noyb’s announcement, they raise questions about how contractors are monitored and how their personal information is handled.
Reuters reported that the complaints involve a platform with more than 900,000 freelance contributors. That figure describes the platform’s reported contributor base, not the number of people actively working or receiving projects at any given moment.
Noyb has asked data-protection authorities to investigate and consider enforcement measures. Filing a complaint is the beginning of a regulatory process. It is not the same as a regulator issuing a ruling or imposing a fine.
What the privacy group alleges
The dispute has three main parts.
Tracking while contractors work
Research associated with Paris-based DiPLab alleges that up to 19 tracking technologies were observed in connection with Outlier work sessions. According to Reuters, the researchers described recordings of interactions such as mouse movements and keystrokes that could allow a session to be reconstructed.
Noyb also questions the disclosure of information to third-party analytics and advertising services. Its concern is not simply whether a platform needs to monitor work quality, but whether the amount and sharing of data are appropriate and lawful.
The published allegations should not be interpreted as proof that every contractor’s screen is continuously recorded or that every third party receives the same information.
Repeated identity and face checks
Contractors may be asked to verify their identity using a government-issued document and a facial image or video through Persona, an identity-verification provider.
Noyb argues that the verification practices are excessive and raises questions about the processing of facial information under European data-protection rules.
Outlier’s own privacy policy acknowledges that the platform and its providers may collect identification documents, selfies, recordings and facial measurements during account checks and at intervals afterward. The policy says verification helps protect the service and reduce fraud.
The existence of that disclosure does not, by itself, settle whether the practices meet all applicable legal requirements.
Requests for personal data
According to the complaints described by Reuters, four workers asked for copies of information held about them between December 2025 and May 2026 and reportedly received no response.
Under the EU’s General Data Protection Regulation (GDPR), people generally have rights to request access to personal information held about them, subject to applicable conditions and exceptions. The complaints ask regulators to examine whether those rights were respected.
Scale AI rejects the allegations
Initial reporting said Scale AI had not immediately responded to requests for comment. However, The Register subsequently published a response from the company.
Scale characterized the report’s claims as “unfounded and inaccurate.” It said Outlier’s practices are disclosed to contributors and require consent, and that the tools involved help verify identities, prevent fraud, protect confidential client information and support accurate payment.
That response is important context. As of October 28, the public reports reviewed for this article do not establish a final regulatory finding on the complaints. Both the organization’s allegations and Scale AI’s position should be considered while the process develops.
Read The Register’s updated coverage for the company’s response.
Why this matters to AI training contractors
Identity verification and activity monitoring aren’t unique to Outlier. Remote contractor platforms may need to confirm eligibility, protect client data and ensure that work meets project requirements. But contributors also need to understand what information is collected, who receives it and how long it is kept.
The case raises wider questions for people who complete AI training and data annotation work: How transparent are the rules? Can workers see the data collected about them? What happens when a platform automatically evaluates their activity or decides whether to offer more work?
These questions are especially relevant when earnings depend on intermittent assignments. As our guide to AI training pay explains, a platform account or an advertised hourly rate does not guarantee ongoing income.
Likewise, being accepted onto a platform doesn’t always mean projects will become available, as discussed in our guide to getting selected for AI training work.
Practical steps contributors can take
If you already work on Outlier or are considering joining an AI training platform, there is no need to assume your account or payments are affected by this news. The complaints do not announce a shutdown, hiring freeze, or change in payout policy.
It is sensible, however, to review a few things:
- Read the privacy policy and verification terms before sharing identity documents or recordings.
- Use official verification channels only. Be wary of unsolicited messages asking for your ID, password or payment details.
- Keep permitted records of your agreements, support requests, and payment history. Do not copy confidential project data.
- Ask about your personal information through the platform’s official privacy contact channels. Your legal rights will depend on where you live and the laws that apply.
- Avoid relying on one source of work. Compare project eligibility, payment terms, and contributor policies, not only advertised hourly rates.
Our guide on finding legitimate remote AI training jobs explains what to check before applying, and our platform comparison covers additional places to research.
For readers who want to review current application requirements themselves, here are the approved links for Outlier, Mercor and micro1.
These are referral links, not endorsements of any platform’s privacy practices or guarantees of available work.
What happens next
European authorities will decide how to handle the complaints, whether further investigation is required, and whether any corrective measures are justified. Those outcomes cannot be assumed in advance.
For now, the confirmed development is that five privacy complaints have been filed, and Scale AI has challenged the allegations.
AI Trainers Club will update this report if regulators publish findings or the company announces relevant policy changes.
Sources: Noyb’s October 27 announcement, Reuters, The Register, and Outlier’s official privacy policy.
Disclosure: This article contains referral links. I may earn a commission when eligible applicants complete qualifying activities, at no additional cost to them. Referral links do not guarantee hiring, assignments, or earnings.
Leave a Reply