Meta keylogs staff typing for AI training — then leaks it
Meta announced a fabulous new programme in April — the Model Capability Initiative! [BBC] a new tool will run on Meta’s computers and internal apps, logging their activity to be used as training data for AI technology. This collected every keystroke, mouse click, and screen image. Meta told the...
Meta announced a fabulous new programme in April — the Model Capability Initiative! [BBC] a new tool will run on Meta’s computers and internal apps, logging their activity to be used as training data for AI technology. This collected every keystroke, mouse click, and screen image. Meta told the BBC it had: safeguards in place to protect sensitive content. Meta employees started a petition in May against the key logging. They asked about these safeguards. There weren’t any: [letter] no completed privacy reviews were provided. The outlined privacy mitigations were vague, and leadership’s confidence in them appeared limited — evidenced by the selective opt-out afforded to executives. The employees were right: [Wired] Meta left potentially sensitive information collected from employee laptops accessible to anyone inside the company. Meta paused the data collection. But it wants to start keylogging again as soon as possible. [Wired] Meta is also sending expensive engineers off to become AI trainers in the Applied AI Engineering Unit! [Reuters] Join or be fired. Applied AI head Maher Saba said: AAI is one of the company’s highest priorities and we’re resourcing it by moving our strongest talent to address it. Therefore, the transfers aren’t optional. The engineers are writing programs to generate training and tests for the AI. One draftee told Wired: [Wired] You have zero purpose in life all of a sudden, you barely interact with anyone, you just have these tasks every week. Just like the task workers who were training the AIs already. Fancy that. Video — PodcastSource: Pivot to AI — Published — Category: Business