Meta Kills Its Employee Spyware After Leaking 45,000 Tables Of Internal Gossip
Nothing screams peak corporate competence quite like building an invasive keylogger to feed AI, only to accidentally leave the entire database wide open for the whole office to read.
Engineers at Meta launched the Model Capability Initiative (MCI) on corporate machines across the United States back in April. The master plan was to record every single mouse twitch, keyboard click, and random desktop screenshot to train artificial intelligence on how real humans use dropdown menus and shortcut keys. Because apparently, asking developers politely how they use a computer was too low-tech for an algorithmic empire.
Staff members immediately revolted, distributing protest flyers in office corridors and collecting signatures against the invasive monitoring. Management tried to pacify the workforce by allowing thirty-minute tracking pauses and formal opt-out requests, but the corporate surveillance engine kept humming along anyway.
The grand experiment finally imploded when misconfigured access control lists exposed 45,000 internal data tables containing keystrokes, personal chats, and performance reviews to company personnel. Chief Technology Officer Andrew Bosworth confirmed the permission blunder, while executives rushed to reassure everyone that outside hackers had not touched the exposed pile of private chatter and screen recordings.
Corporate surveillance always looks brilliant on a slide deck until the digital panopticon points its high-definition lens directly back at the watchers. Turning white-collar staff into unpaid training data guarantees plenty of things, but workplace loyalty and basic database security clearly missed the memo.
Source: Reuters
Comments
Help shape the next version: Add context or suggest a correction. AI review can add points toward a rewrite. Reviews and updates may take time; a full meter does not guarantee a new version.