Meta quietly tests prediction-market app as legal pressure builds on Kalshi and Polymarket
Meta is testing an internal prediction-market application, currently named Takes, that lets employees forecast event outcomes using virtual points rather than real money, according to a Nov. 14 newsletter from The Block. The app awards users points, called “cred” and short for credibility, for correct predictions and deducts them for incorrect ones. The product is not publicly available, and Meta has not decided whether to launch it.
The experiment traces back to June, when the project was first reported under the name Arena. The Block reports that Meta chief executive Mark Zuckerberg instructed employees to develop a standalone prediction-market product, and Takes appears to be the working result of that directive. The app remains an internal test, with no commitment to a public release.
The timing is what makes the move significant. Kalshi and Polymarket, the two most prominent prediction-market platforms in the United States, are facing state lawsuits over whether their event contracts constitute unlawful gambling or regulated financial products. That unresolved question has become one of the sharpest regulatory fault lines in American fintech, and Meta’s points-based design reads as a deliberate attempt to stay on the safe side of it.
By using virtual credibility points rather than wagerable currency, Meta avoids the core legal question that has ensnared its would-be competitors: whether money staked on real-world outcomes is a bet, a derivative, or something else entirely. If the app ever launches publicly, that design choice could allow Meta to build a prediction product at social-network scale without immediately triggering the gambling statutes and commodities laws now being tested in court.
The strategic logic is straightforward. Prediction markets have moved from niche curiosity to mainstream attention, driven by election-cycle volume and heavy media coverage of platforms like Polymarket. A company with billions of users, deep engagement data, and a history of cloning successful formats has an obvious incentive to explore the category. The points-based approach lets Meta learn whether prediction mechanics drive engagement before committing to the regulatory exposure of a real-money product.
Whether users would find a points-based version as compelling as a real-money market is another question. The prediction-market appeal rests partly on financial stakes. But Meta has previously converted social mechanics into engagement engines without direct payouts, and a credibility-scored forecasting layer could sit comfortably alongside existing features. For more on the wider sector, see our DeFi coverage.
DOJ wins approval to seize $1.2 billion stolen in the Binance hack
The U.S. Department of Justice has obtained court approval to seize or restrain more than $1.2 billion in cryptocurrency stolen in the Binance hack, The Block reported. Investigators attributed the theft to TraderTraitor, a group identified as North Korean military hackers. The attackers allegedly stole more than $6 billion from Binance in October, making the seizure a substantial but partial recovery against one of the largest thefts on record.
The recovery effort forms part of a broader campaign that has frozen more than $2.7 billion in allegedly stolen crypto, according to the newsletter, through cooperation between U.S. authorities and private companies. That figure matters beyond this single case. It signals that the public-private tracing model, in which blockchain-analytics firms and exchanges work alongside federal investigators, has reached a scale where even nation-state actors struggle to cash out large stolen positions.
For years, the working assumption in the industry was that stolen crypto was effectively gone once it moved through mixers or cross-chain bridges. The TraderTraitor seizures complicate that assumption. North Korean hacking groups have historically been among the most sophisticated launderers in the space, and the fact that U.S. authorities can now immobilise a fifth or more of a $6 billion haul suggests the tracing advantage has shifted meaningfully toward law enforcement.
The case also underscores the sheer scale of North Korea’s crypto-financing operations. A single alleged theft of more than $6 billion from one exchange dwarfs the annual budgets of many of the country’s known programmes, and the FBI and DOJ have repeatedly linked such thefts to state priorities. The Binance hack, if the attribution holds, would represent a step change even by the standards of a group already blamed for billions in prior exchange and protocol breaches.
For Binance, the seizure is a double-edged development. It confirms the severity of the October incident while demonstrating that a meaningful share of the stolen funds remains traceable and potentially recoverable for affected users. Exchange hack resolutions have historically hinged on how much stolen value investigators can freeze before it dissipates, and a $1.2 billion restraint is a strong early position.
FBI-linked disruption hits FaceStealer, and 16-year-dormant bitcoin moves
Two further items from the newsletter round out a busy week.
First, the DOJ disrupted infrastructure linked to FaceStealer, a North Korean scheme that the FBI says generated more than $400 million in crypto theft over 18 months. The operation sits in the same ecosystem as the TraderTraitor attribution, reinforcing the picture of a coordinated, well-resourced North Korean effort to extract value from the crypto sector through multiple parallel channels. A $400 million take in a year and a half, from a single named scheme, indicates the Binance hack was not an isolated event but part of a sustained campaign.
Second, two bitcoin addresses dormant for more than 16 years moved 100 BTC. The transaction attracted attention because early-era holdings can belong to miners, lost-wallet owners, or original market participants. Movements from wallets created in bitcoin’s first years are closely watched precisely because they are so rare and because their provenance is often unknowable. A 100 BTC transfer is modest in dollar terms by institutional standards, but the age of the coins gives it outsized signalling weight: watchers parse every such movement for hints about whether an early adopter is returning, a lost wallet has been recovered, or an estate is liquidating.
Neither the identity of the holder nor the destination of the coins was disclosed in the report, and speculation about early bitcoin wallets has a poor track record. Still, dormant-coin activity is a recognised market variable, since the potential supply from wallets untouched since 2010 or earlier is one of the lingering unknowns in bitcoin’s supply picture.
What this means for the market
Taken together, the week’s developments sketch two diverging trends.
On one side, enforcement is getting stronger. The $1.2 billion seizure, the $2.7 billion in frozen funds across the broader campaign, and the FaceStealer disruption show U.S. authorities operating with a tracing and seizure capability that would have seemed implausible five years ago. That is a net positive for institutional confidence in crypto custody and exchange security, even as the underlying thefts expose how large the losses remain.
On the other side, the regulatory perimeter for prediction markets is still undefined, and Meta’s cautious entry shows how carefully big tech is stepping around it. If Kalshi and Polymarket lose their state-level legal fights, a points-based Meta product could capture the engagement without the liability. If they win, Meta has a ready-made design it could monetise. Either way, the largest social platform on earth is now formally in the prediction-market game, at least experimentally, and that is likely to pressure regulators to clarify the rules sooner rather than later.
