Kalshi just launched a new AI tool called 'Blanket,' designed to help small businesses hedge operational risks—from weather disruptions to tariff changes—by using its regulated prediction market. The pitch is clean: a third-party AI engine analyzes your exposure, then recommends the perfect event contract to offset it. It sounds like the holy grail of retail risk management. But here's the problem: the mechanism doesn't match the narrative.
Let me state this clearly from the outset: prediction markets are fundamentally ill-suited for the kind of proportional hedging small businesses need. Blanket is a clever UI wrapper around a structural mismatch—a binary payoff structure applied to a continuous risk landscape. This isn't just a liquidity issue; it's a design philosophy flaw.
Context: The Compliance Hack
Kalshi operates under CFTC oversight as a regulated prediction market. Blanket is explicitly not a Kalshi product—it's a third-party tool that 'does not execute trades or handle funds.' This is a deliberate regulatory arbitrage: by positioning Blanket as an independent recommendation engine, Kalshi offloads liability for bad advice while still capturing the transaction fees. The AI analyzes news, weather data, and tariff schedules, then suggests contracts like 'Will the Fed cut rates by 50bps before June?' or 'Will El Niño be declared by August?'
Small businesses, the target audience, are typically retail clients—not qualified investors. They lack the sophistication to understand that these binary contracts don't pay out in proportion to their actual loss. A 10% tariff increase might cost a business $50,000, but the event contract either pays $100 or $0 per share. The 'hedge' becomes a lottery ticket, not a true offset.
Core: The Narrative Mechanism Failure
The core insight here is about narrative decay in prediction markets. The narrative that 'event contracts = insurance' has a half-life of about one severe weather season. Here's why:
First, liquidity concentration. Kalshi's order books are dominated by high-profile events—elections, Fed decisions, Super Bowl winners. The long-tail contracts that small businesses need—like 'Will the price of Midwest corn futures exceed $6.50 by harvest?'—are thinly traded. Blanket may recommend a contract, but if there's no counterparty, the user is stuck. I've seen this pattern before: in 2020, during DeFi Summer, I analyzed Compound's liquidity mining and found that 40% of early liquidity was speculative arbitrage, not long-term holding. The same applies here: the liquidity narrative for long-tail contracts is a hollow yield trap.
Second, basis risk. The contract's payoff is binary, but the business's loss is continuous. If a tariff event happens, the contract pays out a fixed amount, but the actual damage could be three times that. The user is left with a 'hedge' that covers only a fraction of the loss, creating a false sense of security. This is the same gap I identified in 2022 during the FTX collapse—the 'narrative of solvency' blinded investors to the fact that audits and marketing are not the same as reserve verification.
Third, the AI model itself introduces operational risk. Blanket's AI likely uses natural language processing to parse news and weather forecasts. But in tail-risk scenarios—like a sudden tariff escalation or a freak storm—the model may fail precisely when the hedge is needed most. The AI becomes a single point of failure in the very moments of highest uncertainty.
Contrarian: The Real Blind Spot
The contrarian angle is that Blanket's biggest risk isn't from regulators or competitors—it's from the mismatch between the tool's promise and its actual utility. The CFTC may eventually classify event contracts as retail derivatives, forcing stricter suitability rules. But the more immediate threat is reputational: when small businesses realize they can't close their positions or that the payout doesn't cover their losses, they'll complain—loudly. And regulators listen to complaints.
Kalshi's isolation strategy is a double-edged sword. By using a third-party developer, it avoids direct liability for Blanket's recommendations. But if the tool causes concentrated losses, the CFTC will investigate the entire ecosystem. The 'compliance innovation frontier' becomes a regulatory minefield.
The narrative that prediction markets are the next evolution of risk management is a compelling one, but it's built on a foundation of binary logic applied to analog problems. The real evolution is parametric insurance—smart contracts that pay out automatically based on objective data indices. Blanket is a step in that direction, but it's a step that stops halfway.
Takeaway: The Next Narrative Arc
So where does this leave us? The success of Blanket—and by extension, Kalshi's enterprise pivot—hinges on one thing: liquidity in long-tail event contracts. If Kalshi can attract enough market makers to quote tight spreads on weather, tariff, and energy contracts, the tool might work. But that's a chicken-and-egg problem: liquidity follows volume, and volume follows utility. Right now, the utility is unproven.
Watch for the next narrative shift: the convergence of prediction markets with decentralized insurance protocols. If Blanket evolves into a multi-protocol aggregator—recommending not just Kalshi contracts but also parametric insurance policies on-chain—then it becomes a genuine risk management platform. Until then, it's a narrative trap dressed in AI clothing.
