Kaito, the Web3 information platform that specializes in indexing hard-to-reach crypto data, has launched what it describes as the first “verifiable mindshare markets” on Polymarket, opening a new category of prediction markets built on AI-derived sentiment, popularity, and social media chatter.
The partnership also highlights the work of two other partners of Kaito, with the three of them working at the intersection of AI, cryptography, and decentralized markets: Kaito for data and inference, Brevis for zero-knowledge verification, and EigenCloud for auditable AI compute.
Polymarket opens a new category of markets
Kaito announced the collaboration on X, stating that “This is the start of a new category – where anyone, anywhere can predict mindshare, sentiment, and popularity for anything.”
The Kaito-Polymarket product is designed to be verifiable at every stage, which is quite different from traditional sentiment indices or social-media-driven indicators, which rely on opaque algorithms.
Kaito wrote “transforming AI from opaque to a verifiable system is a challenging task,” and to solve that challenge, it partnered with EigenCloud using its EigenAI product, which enables it to turn “what used to be an opaque model into verifiable compute that anyone can audit before Polymarket settles payouts.”
EigenCloud wrote on X, that “AI creates the markets, EigenCloud makes the AI verifiable, and Polymarket brings it to the world,” and went further to describe the launch as “the beginning of verifiable markets.”
Proprietary algorithms will be auditable
The second leg of the technical system comes from Brevis, a zero-knowledge proving service that allows Kaito to keep its proprietary scoring algorithms private while still enabling users to verify that the calculations were performed correctly.
Pointing out the challenge it solves for Kaito and how it ties into the Polymaket, Brevis wrote on X, “prediction markets only work if participants trust the data feeding them. When a company controls the algorithm determining ‘mindshare’ scores, how do you prove those scores weren’t manipulated to influence outcomes?”
Brevis noted the proprietary nature of Kaito’s algorithms. It noted that its role in the tandem is to allow traders to be sure of the proof that the calculations are legitimate without Kaito actually open-sourcing its entire system.
Brevis continued: “We use ZK proofs to verify Kaito’s mindshare calculations ran correctly without revealing the proprietary algorithms themselves, effectively becoming a trustless computation oracle. Traders get cryptographic guarantees, Kaito protects their IP. Combined with EigenCloud, making Kaito’s AI inferences verifiable, the full stack becomes trustless.”
Growing appetite for information markets
The launch allows traders to speculate on real-time measures of online attention rather than on discrete future events, thereby growing Polymarket’s reach at a time when prediction markets are increasingly integrated into mainstream platforms.
Google recently began embedding Polymarket and Kalshi prediction market odds into Google Finance, a move seen by analysts as legitimizing the once-niche market structure.
It also comes at a time when Polymarket consolidates its position as the dominant prediction market platform. The company recently closed a $200 million financing round valuing it at $1 billion, led by Peter Thiel’s Founders Fund, following its accurate predictions during the 2024 US presidential election. The company is reportedly now planning to raise fresh funds at a valuation ranging between $12 billion and $15 billion.
Last month, Intercontinental Exchange (ICE), the parent of the New York Stock Exchange, announced that it would invest up to $2 billion in Polymarket at a valuation of about $8 billion. This has also increased speculation that Polymarket might be IPOing soon.
Kalshi announced a $1 billion raise that saw its valuation jump to $11 billion, highlighting the rise of prediction markets and the increasing popularity of both Kalshi and Polymarket.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.
