Report: AI Labs Now Feeding on Employees' Workplace Conversations
By chasecodewell // 2026-10-16
Copy
 
Artificial intelligence (AI) laboratories are buying workplace records – including Slack messages, emails, documents and source code written by employees – to train AI models, according to a report by Business Insider. The employees whose daily work produces that data are rarely told about the sales and rarely receive payment, the report stated. The data is priced by the terabyte, anonymized and sold to AI companies, according to the report. Firms are receiving six-figure offers that require decisions within 48 hours, the report said, and the market includes data brokers such as Polyshares and Micro1. The AI boom has found what the report described as a convenient source of raw material: other people's working lives. A growing number of defunct and struggling companies are monetizing their internal archives by selling the data to artificial intelligence firms hungry for high-quality training material [1][2].

Brokers Build Market in Operational Data

Polyshares states on its website that it licenses the operating record a business produces and supplies it to AI labs. The company counts as operational data it collects and sells: Slack, email, Drive, spreadsheets, CRM history, support tickets, design files, call recordings, accounting and ERP exports. The firm does not reveal which AI lab receives which data, but cofounder Ryan Lock compared the offers made to companies to those in the early days of the Texas oil boom. "They'll say, 'I'm offering you double market. I need an answer by Friday,'" Lock said, according to the report. Micro1 is another broker that pays companies between $100,000 and $2 million for access to their data, the report said. The company announced it would spend $1 billion on company data over the next 12 months, according to the report. CEO Ali Ansari said he believes that in the future, as much as 10% of U.S. GDP could be spent on training AI models.

Meta and SpaceX Look to Employee Data

Meta tried to launch a program that would have tracked the keystrokes and mouse movements of its employees to train its AI models, according to the report. Employees revolted, and even though CTO Andrew Bosworth said the data collected in that way would be "tightly controlled," it was found to be viewable across the company, the report stated. The data included private conversations and performance records that were supposed to be anonymized. Meta suspended the project. Elon Musk told his employees at SpaceX in August that the company's AI, Grok, would be trained on their data. "In a way, it will be trained on you. It will inherit your thoughts and ideas and beliefs, and I think that's a good thing," Musk said, according to the report. Researchers have warned of a looming "data wall," where the supply of fresh, human-created training data could be completely exhausted [3]. As companies seek new material, AI note-takers are already gathering workplace data for machine learning models, raising privacy concerns as employee communications become training material for artificial intelligence [4].

Stolen Code, Lawsuits and Legal Gray Area

Nick Heiner of Surge AI said he was approached earlier this year by a group offering to sell a large company's codebase for millions of dollars, according to the report. "It became apparent that this group did not have the rights to sell this data," Heiner said. "They had illicitly obtained a copy of the codebase and were just going around to all the data vendors saying, 'Can I get a few million bucks for this?'" Former gig workers for AI training startup Mercor allege in a lawsuit that the company encouraged them to upload work materials without their employers' consent, the report said. Mercor denies the allegation. Emory University law professor Ifeoma Ajunwa said there is little legal recourse for employees because there is no federal privacy law. She called it "an ethically gray area," according to the report. Many jurisdictions require explicit employee consent before collecting or processing personal data for AI training purposes, according to an analysis published by the law firm Michael Best [5].

Spirit Airlines and Other Data Sales

Google sought to buy data from bankrupt Spirit Airlines – including 100 million emails, 500 million Teams chat items, 175,000 personnel records and 1,092,000 time cards, according to the report. The deal has hit a snag, partly because the airline's flight attendants are concerned about how their records will be used. Google told Business Insider it would ensure personal data is either removed or de-identified by an independent third party. In April, OkCupid handed nearly 3 million users' photos to an AI company, and a settlement with the U.S. Federal Trade Commission carried no financial penalty, according to the report. In 2024, a lawsuit against Photobucket alleged it sold users' photos and biometric data to AI trainers without consent. The rush to adopt artificial intelligence is creating new pathways for data breaches, identity theft and corporate espionage, according to NaturalNews.com. AI systems require vast amounts of data to function, but feeding them sensitive corporate or client information is akin to posting confidential files on a public noticeboard, the outlet stated [6].

Conclusion: Employer Signature Is Enough Without Federal Privacy Law

The only party not at the table in the data market is the people who wrote the messages, created the documents and generated the data being sold, the report stated. The anonymization of this data is a promise made by those who sell it and verified by a third party that the buyer pays for, according to the report. Without a federal privacy law, the report added, the only thing needed for employee data to be sold is the employer's signature. Employees who assume that their corporate data is training AI models will become increasingly guarded, and the workplace will be more constrained as a result, according to commentary published by The Economist [7]. Until legal protections are enacted, the market in workplace records will continue to expand, with the terms set entirely by employers.

References

  1. Employee Data Is Being Sold to Train AI Models. https://reclaimthenet.org/employee-data-is-being-sold-to-train-ai-models
  2. Companies Are Selling Workers’ Private Messages and Emails as AI .... https://political.org/2026/04/19/companies-are-selling-workers-private-messages-and-emails-as-ai-training-data/
  3. Are your employees training AI models with company .... https://www.quora.com/Are-your-employees-training-AI-models-with-company-secrets
  4. You’re not sending an email. You’re training a model. https://www.economist.com/business/2026/10/01/youre-not-sending-an-email-youre-training-a-model
  5. AI Training Data and Employee Privacy: Legal Boundaries .... https://www.michaelbest.com/insights/ai-training-data-and-employee-privacy-legal-boundaries-every-business-owner-shou-102kyzg/
  6. The double edged sword How AIs hunger for data makes it cybersecuritys weakest link - NaturalNews.com, November 22, 2025.
  7. If employees assume that their corporate data are training .... https://www.facebook.com/TheEconomist/posts/if-employees-assume-that-their-corporate-data-are-training-ai-models-they-will-b/1627014292790418/

Explainer Infographic

Copy