{"slug": "what-happened-when-meta-tried-to-collect-employee-data", "title": "What happened when Meta tried to collect employee data", "summary": "Meta suspended its Model Capability Initiative (MCI), a mandatory program to collect employee keystrokes and mouse movements for AI training, within two months of announcing it after an employee petition gathered over 1,800 signatures and an internal leak revealed the data was accessible company-wide, Business Insider reported. Meta CTO Andrew Bosworth defended the program internally, saying the data was \"tightly controlled,\" while employees raised concerns it could expose passwords, credit card information, and immigration forms. Meta declined to comment to Fast Company, and Bosworth has said the employees who opposed MCI \"made the entire company worse.", "body_md": "When employees at Meta learned that the company would start [collecting their data](https://www.reuters.com/sustainability/boards-policy-regulation/meta-start-capturing-employee-mouse-movements-keystrokes-ai-training-data-2026-04-21/) to train [AI](https://www.fastcompany.com/section/artificial-intelligence) models, there was a swift backlash. Many of them [openly pushed back](https://www.fastcompany.com/91542004/meta-is-using-mouse-tracking-software-on-employees-now-theyre-pushing-back) on Meta’s plan to capture keystrokes and mouse movements, circulating an online petition and urging their colleagues to sign. \n\nWithin just two months—and after an internal leak—Meta had [put an end](https://www.fastcompany.com/91563851/meta-hits-pause-on-tracking-employee-keystrokes-to-train-ai-after-internal-leak) to the program, at least for now. A [new report](https://www.businessinsider.com/meta-ai-program-prompted-internal-revolt-program-suspended-2026-9) from *Business Insider* chronicles what happened internally as Meta shared its plans with employees, revealing just how quickly the program was shelved over employee concerns and a breach in data privacy. (Meta declined to comment when reached by *Fast Company*.)\n\nThe idea to collect employee data reportedly came from the researchers in Meta’s AI division, who are largely siloed from the rest of the company. One of the researchers who spearheaded the program was Rylan Schaeffer, who reported to the company’s AI chief. When the program—which the company called the Model Capability Initiative (MCI)—was announced, it drew immediate outrage from employees, especially when they learned it was mandatory and nobody could opt out.\n\nAccording to *Business Insider*, when a cybersecurity engineer raised concerns about data privacy and said security seemed like “an afterthought,” Meta CTO Andrew Bosworth went on the defense. “That isn’t remotely a fair characterization! How would any of this end up in ‘the open’? This is tightly controlled data,” Bosworth said, per *Business Insider*. \n\nMany employees were reportedly worried that the program would extract passwords and other sensitive data, like credit card information or immigration forms. Some of them also wondered if MCI would put them out of a job and enable the company to replace employees with AI.\n\nAccording to *Business Insider*, the petition from employees who were opposed to the program eventually got over 1,800 signatures—and workers even [used physical flyers](https://www.fastcompany.com/91542004/meta-is-using-mouse-tracking-software-on-employees-now-theyre-pushing-back) to garner more support and remind their colleagues that protesting their working conditions was protected activity under the National Labor Relations Act. In June, Meta decided people could opt out of the program for up to 30 minutes if they needed to handle anything sensitive, but employees kept pushing for the company to drop MCI altogether. \n\nThe data privacy concerns turned out to be prescient: Toward the end of June, a Meta engineer sounded the alarm after finding that the data gathered by the program was accessible to the entire company—even private conversations. As one employee reportedly put it: “I swear we’re probably the richest yet most incompetent company to exist in tech right now.” Meta suspended the program almost immediately, according to *Business Insider*, and said workers would be able to opt out if it was reintroduced. \n\nThere seem to be conflicting accounts of whether the data Meta collected from employees was used in any meaningful way. Employees told *Business Insider* that one executive claimed the data was not actually used for AI training, while Bosworth reportedly said it proved more useful than anticipated. \n\nIn leaked audio [obtained by *More Perfect Union*](https://www.youtube.com/watch?v=WHhJ-GIF7Lg), Bosworth has maintained that the employees who were openly opposed to MCI “made the entire company worse.” But the blowback Meta faced from its workforce—not to mention the public scrutiny—indicates that the risks far outweigh the benefits of trying to implement a controversial program like this. And as *Fast Company* [has reported](https://www.fastcompany.com/91530650/meta-tracking-employees-ai-training-legal-not-ethical), it invites complex questions about the ethics of collecting employee data, even if it is technically legal to do so on company devices.\n\nThe vocal outcry from Meta employees also points to more public dissent from white-collar employees, who are often not the target of this type of workplace monitoring. In the lead-up to a sweeping round of layoffs in May, many Meta employees expressed their discontent through memes and websites counting down to the announcement, the [*New York Times* reported](https://www.nytimes.com/2026/05/08/technology/meta-ai-employees-miserable.html). \n\nAs the AI boom continues to drive layoffs across the tech industry—and amid growing fears that they could be replaced by AI—some tech workers seem newly emboldened to air their grievances about their working conditions.", "url": "https://wpnews.pro/news/what-happened-when-meta-tried-to-collect-employee-data", "canonical_source": "https://www.fastcompany.com/91606241/what-happened-when-meta-tried-to-collect-employee-data", "published_at": "2026-09-11 20:30:00+00:00", "updated_at": "2026-09-11 21:22:22.348282+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-policy", "ai-ethics"], "entities": ["Meta", "Model Capability Initiative", "Andrew Bosworth", "Rylan Schaeffer", "Business Insider", "Fast Company", "National Labor Relations Act", "More Perfect Union"], "alternates": {"html": "https://wpnews.pro/news/what-happened-when-meta-tried-to-collect-employee-data", "markdown": "https://wpnews.pro/news/what-happened-when-meta-tried-to-collect-employee-data.md", "text": "https://wpnews.pro/news/what-happened-when-meta-tried-to-collect-employee-data.txt", "jsonld": "https://wpnews.pro/news/what-happened-when-meta-tried-to-collect-employee-data.jsonld"}}