Skip to main content

InfoFi Is the New DeFi: How Information Finance Became Web3's $10B Sector in 2026

· 12 min read
Dora Noda
Dora Noda
Software Engineer

In March 2026, prediction markets traded $25.7 billion in a single month. That is more notional volume than most mid-cap equity indices. It is not a bubble, and it is not a meme. It is the clearest signal yet that a new asset class — information itself — has finally found a price.

Welcome to InfoFi.

For years, crypto tried to financialize everything: loans, art, cat pictures, liquidity positions, even carbon. But the one thing markets have always struggled to price — the quality of a prediction, the trust of a person, the value of a dataset — stayed stubbornly analog. That changed in 2026. Three previously separate experiments (prediction markets, on-chain reputation, and AI data marketplaces) converged into a single sector with a single thesis: put skin in the game behind information, and the information gets better.

Wall Street has a name for this thesis. It calls it Information Finance. And on current trajectory, InfoFi will cross $10 billion in sector value before the end of this year.

The Thesis: Information Becomes a Tradeable Asset Class​

InfoFi rests on an uncomfortable idea: the internet was built to distribute information for free, and that business model has failed. Global data volumes are racing from roughly 33 zettabytes in 2018 to a projected 175 zettabytes by 2027. But decision quality — measured by forecast accuracy, content trustworthiness, or model reliability — has not improved at the same pace. If anything, the rise of generative AI has flooded the zone with plausible-looking noise.

InfoFi fixes the incentive problem. Instead of free distribution that rewards virality, InfoFi structures force participants to stake economic capital on what they claim is true. A Polymarket trader who is wrong loses money. A Kaito creator who loses credibility loses reward eligibility. An Ocean Protocol data contributor whose dataset fails validation earns nothing. In every case, the feedback loop is the same: accuracy pays, performance art does not.

This is the same discovery that gave us DeFi in 2020. Back then, the primitive was automated market-making; today, the primitive is price-discovered information. And the early numbers echo the DeFi playbook: fast TVL-equivalent growth, fast fee revenue, fast regulatory attention, and fast copycats.

Pillar 1: Prediction Markets Cross $25B in Monthly Volume​

Polymarket is the headline. As of April 2026, it is reportedly in talks to raise a funding round at a $15 billion valuation — a 66% step-up from its $9 billion round last year. The company has pulled in roughly $400 million, with Intercontinental Exchange (ICE) leading. That is an old-line derivatives exchange operator writing a check into a crypto-native prediction venue. The signal is unmistakable.

But Polymarket is no longer the whole market. Kalshi, the CFTC-regulated competitor, has become a real force on the US retail side. Combined year-to-date volumes across Polymarket and Kalshi have reached roughly $60 billion. A single January 2026 session cleared $701.7 million in one day. Bernstein's research desk has projected $240 billion in full-year 2026 prediction market volume, with a further forecast of $1 trillion by 2030.

The most important structural shift is not volume. It is distribution. Google recently began embedding Polymarket and Kalshi odds into Google Finance, placing prediction market prices next to equity tickers. That is the same move that turned Bloomberg terminals into the default price feed for bonds forty years ago. Once a quote appears inside a default consumer surface, it stops being an alternative source and becomes the source.

Prediction markets are also becoming input data for other financial products. Traders are increasingly using Polymarket probabilities as real-time risk indicators — a kind of crowd-sourced VIX for everything from election outcomes to Fed decisions. That is the "truth machine" thesis: a market-cleared probability is often a better nowcast than a survey, a forecaster, or a pundit.

Pillar 2: Reputation Gets a Price Tag​

If prediction markets price events, reputation protocols price people. The flagship experiment here is Kaito, the AI-powered crypto intelligence platform founded in 2022 by ex-Citadel trader Yu Hu.

Kaito spent 2024 and 2025 building a simple but powerful primitive: YAPS, a mindshare score that ranked crypto creators by AI-measured quality rather than raw engagement. Projects paid Kaito to run Yapper Leaderboards, creators optimized for YAPS, and the platform became the de facto reputation layer for Crypto Twitter.

Then on January 15, 2026, X (formerly Twitter) revoked API access for any app that rewarded users for posting. Kaito shut down YAPS and its leaderboards overnight. Rather than die, the team pivoted. By February 2026, Kaito had relaunched as Kaito Studio — a tier-based, selective creator-brand marketplace with 16 initial partners spanning crypto, finance, and AI. The addressable surface expanded from Crypto Twitter to YouTube and TikTok.

In parallel, Kaito partnered with Polymarket to launch Attention Markets — the first "verifiable mindshare markets" where users can wager on the sentiment and popularity trajectory of brands, tokens, narratives, or public figures. Polymarket supplies the liquidity and the settlement rails. Kaito supplies the ground-truth oracle for mindshare. The merger of the two is the cleanest proof yet that reputation and prediction are not separate sectors — they are two sides of the same InfoFi coin.

Ethos Network is building the same primitive from a different angle. Instead of measuring attention, Ethos directly stakes reputation. Users vouch for other users by locking ETH, write reviews that are weighted by the reviewer's own credibility score, and can propose to slash dishonest actors. The architecture is explicitly modeled on Proof of Stake: the validators are people, the transactions are social, and the slashing is reputational. The result is a "Credibility Score" that other Web3 apps can read as a gating or pricing input.

This matters because Web3's trust problem is now its biggest adoption blocker. If institutions, creators, and mainstream users are ever going to transact on-chain with counterparties they do not know, there has to be a common language for trust. Reputation protocols are the first credible attempt to write that language.

Pillar 3: AI Data Markets Close the Loop​

The third pillar is where InfoFi gets truly structural. AI training requires data, data requires payment, and payment at the scale of modern foundation models requires programmable infrastructure. Ocean Protocol has been building this rail since 2017, and 2026 is the year the rail finally carries freight.

In early 2024, Ocean merged with Fetch.ai and SingularityNET to form the Artificial Superintelligence Alliance (ASI) — a unified token and roadmap around tokenized data, autonomous agents, and decentralized AI. That merger has now reached full operational capacity. In March 2026, Ocean Network launched the beta of its peer-to-peer compute orchestration layer, allowing training jobs to run directly on tokenized datasets without the data ever leaving the contributor's control.

The architecture is clean: Data NFTs represent the underlying intellectual property, ERC-20 DataTokens gate access, and compute-to-data protocols mean model builders can train on private datasets without copying them. A contributor gets paid every time their data is used. A model builder gets training provenance that stands up to EU AI Act audits. An agent economy gets the raw inputs it needs at machine speed.

Adjacent projects are filling out the category. ZENi focuses on privacy-preserving training data using zero-knowledge proofs. LazAI, the Metis-anchored chain that went alpha mainnet in April 2026, is targeting data provenance and inference royalty distribution. Vana is raising data DAOs where users collectively monetize personal datasets. Each of these takes a slightly different cut at the same question: who owns the signal that makes AI valuable, and how do they get paid?

The Flywheel: Why the Three Pillars Need Each Other​

It would be easy to read prediction markets, reputation scores, and data markets as three unrelated bets. That reading misses the flywheel.

AI agents are the connective tissue. An autonomous agent that wants to make money needs three things: a way to generate predictions (data + reasoning), a way to establish creditworthiness with counterparties (reputation), and a venue to monetize its forecasts (prediction markets). Every agent transaction consumes all three InfoFi pillars in a single round trip.

Start with the agent reading a Kaito mindshare feed and an Ocean-tokenized dataset. It forms a thesis and stakes on Polymarket. The outcome resolves. The agent either gains or loses reputation, tracked on a system like Ethos. Its next trade is sized accordingly, and its next data purchase is priced accordingly. Every loop generates fee revenue for the protocols providing each primitive.

Compound that across thousands of agents and millions of human users, and the sector revenue stack starts to look a lot like early DeFi: small per-transaction take rates, enormous aggregate throughput, and a defensible infrastructure layer underneath.

What Could Go Wrong​

Three risks will decide whether InfoFi follows DeFi's arc or stalls halfway.

The first is regulation. Prediction markets sit at the intersection of gambling law, commodities regulation, and election law. Kalshi has CFTC clearance; Polymarket is fighting to replicate that in the US. Any ruling that treats political markets as illegal election interference would carve out the most lucrative category overnight. On the other hand, the ICE investment in Polymarket suggests that the regulated US derivatives establishment is betting the opposite direction.

The second is platform dependency. The Kaito-X API incident exposed the fragility of InfoFi primitives that rely on Web2 data. A single revocation decision on January 15 vaporized YAPS. Any reputation or mindshare protocol that runs on scraped social data is one TOS change from being rebuilt. The survivors will be the ones that migrate to on-chain or cryptographically attested data sources.

The third is the oracle problem in disguise. InfoFi is only as good as its resolvers. A prediction market that resolves wrong loses credibility faster than a lending protocol that gets liquidated. A reputation system that rewards manipulation collapses into a spam economy. Every InfoFi primitive is really an oracle problem with a better UI, and oracle failures have been the single most expensive mistake in DeFi's history.

Building on the InfoFi Stack​

For infrastructure providers, the shape of the InfoFi opportunity is already visible. Prediction markets need ultra-low-latency RPC endpoints so resolvers can ingest real-world events in real time. Reputation protocols need reliable event indexing across hundreds of contract addresses to compute credibility scores. AI data markets need verifiable compute and high-throughput storage. Attention Markets specifically require both mindshare oracle feeds and Polymarket-grade liquidity infrastructure in the same stack.

Generic chain RPC is getting commoditized. The premium is shifting to category-specific data infrastructure — the indexers, oracles, and verifiable compute layers that sit between raw chain state and the InfoFi applications that need to read it.

BlockEden.xyz provides enterprise-grade RPC and indexing infrastructure across Sui, Aptos, Ethereum, Solana, and 15+ other chains — the foundational rails that prediction markets, reputation protocols, and AI data platforms depend on to operate at scale. Explore our API marketplace to build on infrastructure designed for the InfoFi era.

The 2026 Verdict​

DeFi's first act was about making money programmable. InfoFi's first act is about making truth programmable. The two are not the same, but they rhyme. Both start with a narrow primitive that looks like a toy. Both grow into a sector with its own fee curves, its own blue chips, and its own systemic risks. Both eventually reshape the broader financial stack because the old stack cannot replicate their properties at the same cost.

The numbers today already justify the comparison. Twenty-five billion in monthly prediction volume. A fifteen-billion-dollar Polymarket valuation. An $8.5 billion single-month print in January. Combined year-to-date Polymarket-plus-Kalshi volumes pushing $60 billion. Kaito Studio partnerships across sixteen brands in its first month. An ASI-aligned data economy that now has a live peer-to-peer compute layer.

If even half of Bernstein's $240 billion 2026 prediction-market forecast lands, InfoFi will be the first post-DeFi category to generate a verifiable ten-figure revenue stack. The rails are being laid right now. The only remaining question is who owns the toll booths when the traffic arrives.

Sources​

Share on Twitter