Sift

Anthropic sells Claude Opus input tokens for between $4 to $5 per million, depending on model version. On Taobao, resellers move the same access for a fraction of that price, sometimes as little as 10% of the list price. That gap doesn’t close itself. Something is funding it, and it isn’t one thing. It’s several kinds of fraud, stacked, each one hitting a different point in the customer journey. Tokens are behaving like a currency, and fungible currencies attract thieves and counterfeiters.

Over time, Trust and Safety teams deal with a problem where the more fraud that’s caught, the more manual review queues grow. Technically, they can remedy this issue by cutting review volume to protect the customer experience, but when they do this, false declines will creep up instead. Recent industry data suggests this doesn’t need to be a trade-off, as the programs that are closing this gap are instead relying on better decisioning rather than bigger queues.

Responsible gambling and fraud prevention used to live in separate departments with separate tools and separate KPIs. That split does not hold up anymore. The same behavioral signals that flag a fraudster opening a fake account can also flag a player heading toward harm, and iGaming operators that connect these two disciplines are catching more of both. A June 2026 study found that fraud rates in iGaming rose nearly 40% overall since 2024 with an 18% increase year-on-year.

Two-sided platforms carry a much heavier fraud burden compared to single-sided retailers. Every buyer, seller, renter, and payer is a potential attack surface, and one bad actor can hit both sides of a transaction at once. For commerce marketplaces and peer-to-peer (P2P) platforms, fraud protection functions as trust infrastructure, determining whether people keep transacting on the platform at all.

While every operator budgets for promotions as a customer acquisition cost, very few budget for the version of that cost that never converts into a real player. Bonus abuse and multi-accounting now account for the single largest fraud category in iGaming, making up 64% of fraud according to a recent study. But unlike chargebacks or payment fraud, this kind of loss hides inside your own marketing spend until it shows up as a margin problem nobody can explain.

In just Q1 2026, U.S. iGaming has generated about $3.04 billion, marking over 20% growth year-over-year and nearly $1 billion in April 2026 alone. For the full 2025 year, iGaming hit over $10 billion in revenue, up 27.6% over the previous year. With it being such a profitable industry, there’s no wonder why it’s ripe with fraud.

Missed Part 1? Read it here. Prediction markets make a bold claim: that they are engines of truth discovery. Proponents argue that by aggregating collective judgment, prediction markets generate forecasts that are far more valuable than traditional gambling or sports betting. That claim has merit. But a prediction market’s output is not merely a well of raw, collective intelligence.

When a social media profile features photos, connections, and a history behind it, it looks like proof a real person is behind it. Fraudsters know this, and they build that appearance on purpose. For fraud teams, treating social presence as a trust signal is one of the fastest ways to approve a transaction that should have never gone through. Most tools that pull in social data are answering a narrow question: does this email, phone number, or name show up somewhere else online?

I recently joined Jerry Hoff, CEO of AppSec Training, for a Blueprint Series session on fraud signal sharing, and it’s a topic I keep coming back to. Fraud, trust and safety, and security teams often work from separate systems with no shared view of the same bad actor. That gap slows response time and lets repeat offenders move across teams undetected. Here’s what I’ve learned building signal-sharing programs across internal teams and vendors.

This April, U.S. federal prosecutors charged an Army sergeant with using sensitive classified information to bet on Polymarket that U.S. ​forces would enter Venezuela and remove Maduro from power. About the same time, Kalshi disclosed that it had fined and suspended three congressional candidates for five years after they traded on prediction markets tied to their own candidacies.