Meta pauses program to track divisive employees after an internal security issue revealed potentially sensitive data collected through the initiative to other employees.
“We have carefully designed this program with privacy safeguards in mind, and while there is no indication at this time that any data was improperly accessed by Meta employees, we are pausing it while we investigate,” says company spokeswoman Tracy Clayton.
In April, Meta made the Model Compatibility Initiative (MCI) tool available to U.S. employees. According to employees who filed petitions against it over privacy, security and personal liberty concerns, the tool “collects computer input such as mouse movements, click and keystroke locations, as well as screen content.” After MCI was launched, employees could not resign, but this changed to a constrained extent after employee protests.
Meta executives have repeatedly defended the data collection project, saying it is necessary to train artificial intelligence systems to run computer software in the same way humans do and that workers are the best examples of artificial intelligence to learn from.
On Monday, Meta Engineer issued an internal security advisory stating that databases containing information collected by MCI had been leaked to company insiders.
A former employee actively involved in opposing MCI describes the deviation as a “mess” – one that employees expected. “When employees raised concerns, management doubled down and failed to address the risks employees raised regarding the security and privacy of employee and customer data,” this person says. “Leadership has clearly created an authoritarian environment where employees are no longer respected or heard.”
But after critical comments expressing frustration over security issues appeared on internal forums on Monday, Meta shocked some of its employees by pausing MCI altogether. She informed WIRED of the development before announcing it to employees, according to two people familiar with the matter.