Meta AI launches private Incognito Chat amid scam-ad scrutiny and layoff reports

Incognito Chat is Meta AI's privacy-flavored response to the rising consumer concern that chat history is a liability — both for the user (data retention) and for Meta (legal exposure). The mode promises ephemeral conversations not retained in user history, though as with any privacy claim from Meta the implementation details and retention policies will matter more than the marketing.

Mechanically the feature follows ChatGPT's similar temporary-chat mode and reflects a broader industry recognition that always-on memory has downsides. Meta's twist is the platform context — these conversations happen inside Instagram, WhatsApp, and Messenger, where Meta has separate data agreements with users that complicate the trust story.

Competitive context: the launch lands in a tough week for Meta. Fox News reporting flagged ongoing concerns about Medicare scam ads targeting seniors via the platform, and broader reports suggest Meta is weighing a ~10% layoff. Yann LeCun spent the weekend on a long French-language podcast and posted about world models — and separately Hiro Capital announced it co-led a seed round in LeCun's new Advanced Machine Intelligence (AMI) venture alongside Greycroft, Cathay, NVIDIA, and Bezos Expeditions, fueling speculation about his Meta tenure.

What to watch: Incognito Chat's actual retention behavior (independent audit, ToS language), whether layoff reports materialize and what they reveal about Meta's AI infrastructure ROI, and whether LeCun's AMI grows from a side-project advisory role into something more disruptive to Meta's research agenda.