What happened
Meta is piloting a prediction game aimed at its own workforce, according to a NewsBTC report published Monday. Employees forecast outcomes on a set of questions and earn or lose internal points based on how calibrated they are. There is no real-money stake, no public sign-up, and no indication the product is being prepared for external users in its current form. The report frames it as an internal product test, the kind Meta runs dozens of in any given quarter inside its Reality Labs and central product groups.
The mechanics look familiar to anyone who has used Kalshi or Polymarket. Users pick a side on a binary question, the implied probability moves with volume, and the house tallies results after resolution. What is different is the unit of account. Points, not dollars. That one change carries most of the legal and regulatory weight in this story.
Why it matters
A points-only prediction game is not a prediction market in the US regulatory sense. The CFTC's long-running fight with Kalshi over election contracts, and its earlier scrutiny of Polymarket, both turn on real-money event contracts. Strip out the cash, and you are running an internal forecasting tournament, closer in spirit to the Good Judgment Project than to a regulated exchange. Meta appears to have read that line carefully.
That matters because the prediction category is finally having a moment. Polymarket cleared more than $3 billion in cumulative volume during the 2024 US election cycle and has kept a steady base of political and sports markets since. Kalshi won its court fight to list election contracts and has pushed into sports event contracts through 2026. A FAANG name running internal trials in the same category is a signal that the big platforms are watching, even if they are not ready to touch the regulated version yet.
