Meta Is 'Pausing' Employee Tracking Program After It Let the Whole Company See Sensitive Data

My take: This one bugs me and let me tell you why. Meta launched an internal program in April called the Model Capability Initiative that logged its own employees' keystrokes, mouse movements, and conversations to train AI models. In June, that private data (conversations, performance metrics) was accidentally exposed to the entire company. Meta says it "paused" the program and found no evidence of improper access, but the detail that sounds ugly to me is that no one knew the program even existed until it leaked. So they were watching you and you had no clue. That the same companies building surveillance AI to sell to clients test it first on their own team is a pattern worth watching closely 🙄. AI is an incredible tool, but who uses it and for what matters enormously, and here we should demand transparency. Sure, that is my opinion.