Prediction markets need rethink, Buterin says

In an X post Buterin argued that prediction markets have “over-converged” toward “unhealthy” product-market fit centered on near-term price betting and high-engagement gambling-style markets, rather than long-horizon uses that generate durable informational value and practical risk management. He said onchain prediction markets paired with large-language models could be repurposed into “general hedging mechanisms” aimed at stabilizing purchasing power for day-to-day goods and services.

Buterin’s sketch centers on building price indices across major spending categories and regions and listing prediction markets on each category, then tailoring hedges to individuals and firms. He described a setup in which each user runs a local AI model that understands their spending patterns and constructs a personalized basket of prediction market shares representing “N” days of expected future expenses. The basket would function as a targeted offset if the cost of living rises, while users could still hold other assets for growth, according to his outline.

The critique lands as prediction markets expand rapidly in the U.S. and beyond, especially around sports. During Super Bowl LX, prediction-market trading volumes surged, with reported activity spread across platforms including Kalshi and Polymarket, reflecting the way event contracts can be traded in and out before resolution and generate high turnover. That growth has pulled prediction markets closer to mainstream sports and finance conversations, including high-profile partnerships and investments tied to sports figures.

Buterin’s concern is that the same growth dynamic is pushing platforms toward the most viral, short-duration markets at the expense of longer-term hedging products. He is effectively calling for prediction markets to compete less with sportsbooks and short-dated crypto price bets, and more with instruments that let households and businesses reduce exposure to volatile real-world costs.

Regulation is also shifting at a moment when product direction is being debated. In early February, the U.S. Commodity Futures Trading Commission withdrew a Biden-era proposal that would have prohibited sports- and politics-related event contracts, which have become a major part of prediction-market activity. The withdrawals also included related prior guidance that had warned firms away from sports-related event contracts, according to reporting on the agency’s actions.

That policy change has intensified scrutiny from lawmakers and market participants over how far prediction markets should extend into sensitive categories and how much oversight is appropriate as volumes rise. A recent report described senators urging regulators to tread carefully while noting the CFTC’s faster pivot after leadership changes and the formation of advisory efforts involving industry executives.

Against that backdrop, Buterin’s proposal is a different kind of expansion argument: not “more markets for more events,” but a redirection toward hedging real economic exposure. The concept would move prediction markets closer to a consumer-finance utility, where the primary user goal is risk reduction—covering future costs for rent, groceries, energy, healthcare, or other region-specific spending baskets—rather than attempting to profit from binary outcomes.

The open question is whether markets built on granular spending indices can attract enough liquidity to offer reliable pricing and tight spreads, and whether users will accept the data and privacy tradeoffs implied by personalized hedging—even if Buterin’s design leans on local AI models rather than centralized data collection. His outline explicitly relies on local models understanding expenses and assembling the basket, a choice he framed as part of making the system usable and privacy-preserving.

For now, Buterin’s message is that prediction markets have proven they can scale, but the industry’s next step should prioritize products that hedge real-world price exposure. His warning arrives as U.S. regulators ease off a proposed crackdown on the very categories that have driven much of the sector’s recent growth, leaving platforms with more room to decide what they want prediction markets to become.