The European privacy organization noyb, together with the research groups Intérêt à Agir, DIPLab and Reversing.Works, has filed five complaints against Scale AI under the EU's General Data Protection Regulation, alleging that the company's Outlier platform subjects the gig workers who train AI models to invasive and unlawful surveillance. The complaints were filed on Monday with data protection authorities in France, Italy, the Netherlands and Austria.
Scale AI's Outlier platform employs more than 900,000 freelance workers who perform the human labor behind AI training — writing prompts, ranking model answers and annotating data. The platform's customer list reads like a who's-who of the AI industry: Meta, OpenAI, xAI, Google and Anthropic, along with the US government. Reuters reported the complaints on Monday morning. For more context on this story, see our ongoing more AI stories.
Trackers From TikTok, LinkedIn and Facebook on Workers' Devices
The technical analysis underpinning the complaints found that while workers are on the Outlier platform, tracking code from TikTok, LinkedIn, Facebook, the survey platform Sprig and the data analytics company Datadog is placed on their devices. According to noyb, these trackers send workers' personal data to third parties every few seconds, including IP addresses, contact details, unique identifiers and information about the page being visited.
"Such constant and extensive tracking while working really goes against all reasonable expectations," said Lisa Steinfeld Leiser, a data protection lawyer at noyb. "Why should LinkedIn or Facebook know which tasks you're working on?"
The group argues the practice violates the GDPR's requirements around purpose limitation and data minimization, and that workers — classified as freelancers rather than employees — have no meaningful way to refuse the tracking without losing access to income.
Forced Biometric Verification Through Persona
The second pillar of the complaints concerns identity verification. Outlier requires workers to repeatedly pass biometric verification carried out by the identity company Persona, including video recordings of their faces and uploads of government-issued ID documents. Scale AI says the process exists to prevent identity fraud and verify workers' skills.
noyb argues the practice goes far beyond what is necessary, and points to Persona's security history: in October 2025, approximately 70,000 photographs of government IDs stored with the company were exposed by hackers. The complaints also note that Persona reserves broad rights to collect personal data, including IP addresses, device information, geolocation, addresses and consumer credit data.
"Platform work seems to be fertile ground for a whole new type of workplace surveillance," said Thomas Le Bonniec, a sociology researcher at DIPLab. "This would be seen as unacceptable in any other form of employment in the European Union. This system only thrives because of its deliberate lack of transparency."
The substance of the complaints is worth spelling out, because it describes a working environment most people would never accept in a conventional job. A worker annotating data or rating model outputs on Outlier is, according to the technical analysis, simultaneously feeding advertising platforms with a stream of behavioral signals — the pages they visit, identifiers that link those visits to their broader online profiles, and details that can locate them. noyb notes that this data can then be used for targeted advertising and profiling, purposes that have nothing to do with paying a worker for completed tasks.
The complaints also target the asymmetry of the arrangement. Workers are classified as freelancers, which places them outside most labor protections, yet they are subject to controls — continuous monitoring, repeated identity checks — that resemble the most intensive forms of employee surveillance. Companies in the platform economy classify workers as self-employed, which often excludes them from social benefits and affects their wages, and the complaints argue that data protection violations compound that vulnerability rather than existing separately from it.
The Hidden Workforce Behind Frontier Models
The complaints arrive amid growing scrutiny of the labor that powers AI systems. More than 40 million people worldwide now have their work organized through digital platforms, and AI annotation has become one of the fastest-growing segments. The data labelers and raters who make chatbots like ChatGPT, Gemini and Grok polished and safe typically work in isolation, paid per task, with their output feeding directly into models whose valuations now run into the hundreds of billions.
"The violations highlighted in the complaints against Scale AI clearly show that the AI boom relies on a vast workforce whose privacy is at stake," said Baptiste Delmas, project coordinator at Intérêt à Agir. "Data protection authorities must now take these issues seriously and impose strict enforcement measures."
The US government's presence on Outlier's customer list adds a further dimension. Scale AI has won significant federal contracts, and the complaints highlight how the same platform that supports government and frontier-lab work is, according to the filers, operating outside European privacy law.
What Happens Next
Under the GDPR, the French, Italian, Dutch and Austrian authorities that receive the complaints can order Scale AI to stop the practices, impose fines of up to 4 percent of global annual revenue, or both. Enforcement against US platform companies has historically been slow, but noyb has a track record of following complaints through to binding decisions, including landmark actions against Meta and Google over forced consent.
For the AI industry, the case is a test of whether the labor layer that makes models usable will be held to the same privacy standards as the products themselves. If the authorities side with the complainants, AI training platforms operating in Europe could be forced to strip third-party trackers from their work environments and rethink biometric identity checks — changes that would ripple through an industry that depends on exactly this kind of distributed, flexible workforce.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
