Valge

The growth numbers are hard to ignore. Monthly volume went from under $100 million in early 2024 to over $13 billion by the end of 2025. Active users went from around 4,000 to over 600,000 in the same period. That kind of trajectory doesn't happen in a niche; it happens in a category that's finding mainstream traction. How they actually work The core mechanic is different from traditional betting. Users don't wager against a house; they trade probability contracts with each other. You buy a YES or NO position on an outcome, and you can exit that position before the event resolves. The platform takes a fee on matched trades and has no exposure to the result. It's structurally closer to a futures exchange than a sportsbook. What people are betting on Elections and political events are the obvious use case, but the markets have expanded well beyond that. Active contracts currently exist on AI product releases, entertainment awards, crypto price movements, geopolitical events, and considerably more niche outcomes. The range is partly what makes these platforms interesting from a market design perspective. Price discovery works on almost anything where there's genuine uncertainty and enough participants. The main platforms Kalshi is US-regulated and fought a legal battle to offer election contracts, which a federal appeals court cleared in 2024. Polymarket operates on-chain and has attracted significant volume in crypto-native audiences. PredictIt has been running under a CFTC no-action letter for years and remains one of the oldest active platforms. On the institutional side, DraftKings and FanDuel are developing prediction products, with FanDuel working alongside CME Group. Robinhood and Interactive Brokers are also reportedly exploring the space. Why is this happening now Regulatory clarity, particularly in the US, is the biggest unlock. Beyond that, the post-COVID retail trading wave brought a large new audience to probabilistic thinking about markets. There's also a structural shift in how younger demographics relate to financial products — peer-driven, transparent, and sceptical of traditional intermediaries. Where it goes Some projections put annual volume at $1 trillion by the end of the decade. Whether that materialises depends heavily on how regulators in other jurisdictions respond, and whether platforms can manage the reputational risk that comes with politically sensitive markets. How are regulators in your region treating prediction markets, and do you think the current frameworks are equipped to handle something that sits this awkwardly between gambling and finance?

I think those concerns are understandable. Anytime a new technology enters an industry, people worry about how it might be used. But in most regulated markets the core mechanics of slots (RNG, RTP, volatility models) still have to be tested and certified, regardless of whether AI is involved in development or not. So AI can change how games are designed or personalised, but it usually can’t change the underlying math once the game is approved. Where AI is probably going to have the biggest impact is behind the scenes — things like fraud detection, player behaviour analysis, support automation, and marketing. Player-facing features like AI-generated slots are interesting experiments, but they’re still very early. The bigger question for me is how regulators will approach it. Will they treat AI-generated games the same way as traditional ones, or introduce new certification rules? That might end up shaping the market more than the technology itself.

Hello to all of you. Since AI is used everywhere these days, I'd want to discuss what I've seen in the iGaming industry and how it was applied there. iGaming has always been quick to adopt new technology. If a tool helps increase engagement, retention, or revenue, operators usually test it early. AI is no exception. What started a few years ago as a back-office experiment is now becoming part of the player experience itself. Until recently, most AI in iGaming worked quietly behind the scenes. It helped operators personalise bonuses, detect fraud, analyse player behaviour, and automate customer support. Players rarely noticed it, but it improved efficiency and helped platforms manage risk. What changed in the last year is that AI has started appearing directly inside the product. One of the most talked-about examples is SlotGPT, launched by Stake at the end of 2025. The idea is simple - players can generate their own slot games using a short prompt and then play them. The AI creates the visual theme and structure, while core parameters like RTP and maximum win remain fixed. At first, some people framed this as the "end of game providers." In reality, that’s unlikely. The value of a successful slot is not just graphics or a theme. It’s the underlying math model, volatility balance, feature design, and years of iteration based on player behaviour. AI can generate a shell quickly, but it doesn’t replace the experience of studios that optimise games for retention and long-term engagement. Where AI is already proving useful is in areas like personalisation, fraud detection, player support, and marketing automation. These tools help operators understand player behaviour faster and respond in real time. So the real shift isn’t that AI will replace the industry overnight. It’s that AI is slowly moving from an internal optimisation tool to something players interact with directly. And that raises an interesting question for the next few years: Will AI mostly remain a background optimisation layer, or will it start shaping entirely new types of gaming experiences?

Fair take. I’d just add one nuance: “lighter-touch” doesn’t automatically mean “no rules” or “no outcomes”, it usually means the enforcement style is different and not always as visible/public as the big European regulators. From a player’s point of view, I’d treat any offshore licence as a baseline signal, not a safety guarantee. The better indicators are more practical: how long the brand has been around under the same ownership, whether withdrawal limits and timelines are clearly stated, whether they do KYC early, whether games are from reputable providers with certified RTP, and how they handle complaints when something goes wrong.

We’ve all been hearing how “hot” Anjouan has become in the licensing world, so here are a few fresh datapoints from the register that put the hype into numbers. Key statistics (as of 3 March 2026) The statistics show 1,209 active operators. What stands out most is how renewal-heavy 2026 will be: 952 licences, or 78.74% of all active licences, expire this year. So 2026 isn’t just another growth year. It’s a real stress test for renewals. Issuance momentum The growth curve explains the expiry spike. Issuance jumped from 44 licences in 2023 to 430 in 2024, then 582 in 2025. With a roughly annual cycle, that 2025 cohort rolls straight into 2026 expiries: 579 out of 582 expire in 2026. The 2024 cohort also feeds into it heavily, with 344 expiring in 2026. B2B vs B2C The 2026 expiries are overwhelmingly B2C: 90.14%. That’s important because B2C is where real-world friction usually shows up first, things like PSP appetite, KYC/AML expectations, player-facing trust, and dispute handling. If there’s going to be a visible “quality filter” in renewals, it’s most likely to appear here rather than in quieter B2B activity. What’s next The core question for 2026 isn’t only “are licences being issued?”, but “how many operators renew once the first 12 months are over?” If renewal rates hold, the active licence count can keep rising even if new issuance slows. If renewals are weak, you’ll still see plenty of movement, but more churn than real growth. Questions for the audience Do you think 2026 is the year Anjouan proves staying power, or the year we start seeing the ceiling? What renewal rates would you expect once the big 2025 cohort comes due? And in your experience, what usually drives non-renewal the most: payments/PSPs, compliance burden, reputational concerns, or operators simply rotating to the next “fast and cheap” option? Anjouan – Risk dashboard.pdf

Interesting topic, thanks for sharing the Cft letter. I don’t think the investigation alone destroys Curaçao’s credibility – if anything, outside scrutiny can be a good thing. The real test will be what CGA does next. If this ends with no visible changes to oversight or complaints handling, it will just confirm the old reputation and push more operators and players toward other jurisdictions. But if they use the pressure to tighten checks and become more transparent, Curaçao 2.0 could still recover some trust. Right now it feels like a turning point rather than a verdict. Curious to see which way they go.

Hi and welcome! Good question. In most Anjouan setups, it works like this: The licence is issued to one company -->That company then runs many brands/domains under the same licence. So it’s usually one operator with lots of skins, not a Curacao-style “master licence renting to many separate companies”. If you open the T&Cs on those sites, you’ll normally see the same legal entity repeated in the footer – that’s the real operator behind all the brands. If you want to dig deeper into how these structures work, feel free to DM me.

Hey all. New around here and looking forward to exploring the forum, reading stories and maybe sharing some of my own later on.