According to a BBC report, Meta has paused a company-wide program that tracked employees' computer usage to gather data for training artificial intelligence models, following internal backlash and privacy concerns.
The Model Capability Initiative
The program, named internally the Model Capability Initiative (MCI), was started only two months ago as part of an effort by Meta to collect data on how people used computers, including mouse clicks and keystrokes, that could be used to train AI models. A Meta spokesman confirmed to the BBC that MCI was "on pause for now" as the company investigates. The pause came after the company realized some of the collected data had been left potentially accessible to anyone inside Meta. "We have no indication at this time that any data was improperly accessed by Meta employees," the spokesman added.
Internal Frustration and Backlash
The initiative was met immediately with upset from employees who were to have their every online action at work tracked and recorded. Frustration culminated in a petition signed by nearly 2,000 Meta workers demanding that the MCI program be cancelled. In an initial response, Meta said it would allow workers to not be tracked for up to 30 minutes at a time. One current employee told the BBC that was "just an attempt at damage control." Another employee said the tracking "was forced on us, there was no consent," adding, "I've never seen morale here so bad." The employee also noted that while many technical workers are open to improving Meta's AI models and competing with Anthropic and OpenAI, the forced nature of the program angered staff.
"The direction this company is going in is depressing. Exhausting and depressing." — A former Meta employee, quoted by the BBC.
Implications for Enterprise AI and Data Privacy
This episode highlights the tension between AI development and employee privacy, a challenge relevant to enterprise technology leaders. Meta is spending up to $145bn (£109bn) on AI initiatives this year alone, and has conducted extensive layoffs and reorganizations around AI. The tracking program was intended to give Meta a competitive edge in AI, but the privacy breach and backlash demonstrate the risks of aggressive data collection. The table below summarises the key aspects of the MCI program and its fallout:
| Aspect | Details |
|---|---|
| Program name | Model Capability Initiative (MCI) |
| Duration started | Two months prior to halt |
| Data collected | Mouse clicks, keystrokes, online actions |
| Halt date | Monday (date not specified) |
| Reason for halt | Data left potentially accessible to all employees |
| Employee petition | Nearly 2,000 signatures demanding cancellation |
| Compromise offered | 30-minute opt-out periods |
| Company stance | "No indication of improper access" |
| AI spending this year | Up to $145bn |
For CTOs and digital transformation leaders considering employee monitoring for AI training, the Meta case underscores the importance of consent, transparency, and data security. The backlash from technical staff — who are otherwise supportive of AI progress — reveals that even within tech companies, privacy fears can derail such initiatives. The pause is a reminder that data governance must be built into AI training programs from the start, not as an afterthought.
Sources:
