InfoFi's Trial by Fire: How Tokenized Attention Survived X's Ban and Found Its Real Purpose

On January 15, 2026, Nikita Bier — X's head of product — posted a single announcement that erased hundreds of millions of dollars from a nascent crypto sector overnight. X would immediately revoke API access for any application that financially rewarded users for posting. Within 24 hours, KAITO plunged 17.7% and COOKIE cratered 15.5%. The InfoFi sector's total market cap dropped 13%, falling from roughly $367 million to $359 million.

The "attention economy" experiment that Vitalik Buterin had envisioned just fourteen months earlier seemed dead on arrival. But what happened next tells a far more interesting story — one about what survives when the easy money disappears.

From Prediction Markets to Information Finance: Vitalik's Original Vision

The intellectual roots of InfoFi trace back to November 2024, when Vitalik Buterin published "From Prediction Markets to Info Finance." Fresh off Polymarket's breakout during the US presidential election — where the platform's odds proved more accurate than traditional polling — Buterin argued that prediction markets were just the beginning of something larger.

His thesis was deceptively simple: if markets can price the probability of future events, they can also price the quality and credibility of information itself. Buterin described "info finance" as a fair and democratic system where tokenized mechanisms could govern everything from DAO decision-making to scientific peer review.

Crucially, he noted that prediction market prices are "bounded between 0 and 1," making them structurally resistant to the reflexivity, pump-and-dumps, and greater-fool dynamics that plague traditional crypto speculation.

The essay ignited a builder frenzy. Within months, a new category emerged: InfoFi — Information Finance — protocols that aimed to turn attention, reputation, and knowledge into tradable on-chain assets. Three projects raced to define the space.

The Three Pioneers: Kaito, Intuition, and Cookie DAO

Kaito: Wall Street Meets the Attention Economy

Kaito's origin story reads like a crypto founder archetype turned on its head. Yu Hu, who spent nearly a decade as a trader at Citadel — one of the world's most sophisticated hedge funds — founded Kaito in 2021 after growing frustrated with crypto's chaotic information landscape. While Wall Street had Bloomberg terminals centralizing data, crypto's signal was scattered across X, Discord, Telegram, governance forums, and podcasts.

Kaito's first product was a vertical search engine indexing these fragmented sources through AI-powered semantic analysis, offered as an $833/month subscription to traders, funds, and AI agent companies. But the real breakthrough — and eventual controversy — came with "Yaps," a system that quantified social media attention into tokenized scores. Users earned Yaps by posting quality crypto content on X, creating a direct financial incentive for information production.

At its peak, the Yaps system turned X's crypto community into a buzzing hive of content creation. The problem was that it also turned it into a spam factory.

Intuition: Building the Internet's Trust Layer

While Kaito gamified attention, Intuition tackled a deeper problem: how do you verify what's true in a decentralized world?

Backed by $8.5 million in funding from ConsenSys and other investors, Intuition launched the world's first token-curated knowledge graph as a Base L3 built on the Arbitrum Orbit stack. The architecture introduces two novel primitives:

• Atoms — token-curated identifiers for any entity (a person, project, claim, or data point)
• Triples — structured subject-predicate-object assertions that form a decentralized knowledge graph

Think of it as Wikipedia, but where every claim has a financial stake behind it. Network participants create, curate, and stake TRUST tokens on identities and claims, earning rewards for accuracy and losing stake for misinformation. By March 2026, the protocol had recorded more than 4.3 million attestations from approximately 220,000 accounts — modest by DeFi standards, but significant for a system that's essentially building a machine-readable layer of verified truth.

The implications extend far beyond human users. As AI agents proliferate — there are now over 250,000 daily active on-chain agents — they need trustworthy information feeds to make financial decisions. Intuition positions itself as that feed.

Cookie DAO: The AI Agent Index

Cookie DAO carved out a third angle: rather than incentivizing content creation (Kaito) or verifying truth (Intuition), it built the data layer for tracking AI agents themselves.

Through cookie.fun, the platform aggregates over 7TB of live on-chain and social data to track 1,500+ AI agents across blockchain ecosystems. The dashboard measures three critical dimensions: mindshare (the volume of attention a project receives), sentiment (whether that attention is positive or negative), and top voices (the key opinion leaders driving narratives).

In the AI agent era, where autonomous systems manage billions in DeFi capital, knowing which agents are gaining traction — and which are losing credibility — is itself a form of alpha. Cookie DAO aimed to become the Bloomberg Terminal for the machine economy.

The January Crisis: When X Pulled the Plug

The InfoFi thesis hit a wall on January 15, 2026, when Nikita Bier announced X's new policy. The platform would immediately revoke API access for apps rewarding users for posting, citing "a tremendous amount of AI slop & reply spam."

The ban was surgical and devastating. Kaito's entire Yaps incentive model depended on X's API to track and reward engagement. Cookie DAO's analytics similarly relied on X data feeds. Overnight, the infrastructure foundation of the two largest InfoFi protocols was yanked away.

The market reaction was brutal. KAITO fell 17.7% to $0.57. COOKIE dropped 15.5% to $0.038. The broader InfoFi category shed 13% of its market cap. Crypto commentators declared the sector dead.

But the X ban revealed something important: it separated the projects that were genuinely building information infrastructure from those that were simply arbitraging social media engagement for token rewards.

The Pivot: From Paying for Posts to Pricing Credibility

What happened after January 15 is where the InfoFi story gets interesting.

Kaito moved fastest. Within days, it sunset the open Yaps reward system entirely and launched Kaito Studio — a selective, tier-based platform where brands and creators commission targeted marketing campaigns. Instead of rewarding anyone who posts, Kaito now curates a network of verified creators with proven audience quality. In early March, the platform partnered with Polymarket to launch Attention Markets — prediction markets for social trends that merge Kaito's attention data with Polymarket's proven market mechanics.

Yu Hu signaled personal conviction by purchasing 1 million KAITO tokens (approximately $1.4 million) with personal funds and staking all of them, becoming the second-largest staker on-chain.

Cookie DAO announced a transformation into a prediction market platform in February 2026. Rather than simply indexing AI agents, Cookie now enables users to bet on which agents, narratives, and projects will capture mindshare — turning passive data consumption into active information pricing.

Intuition, notably, was least affected by the X ban. Because its knowledge graph doesn't depend on any single platform's API — it operates as its own L3 chain with native staking mechanics — the protocol continued growing through the crisis. This architectural independence proved prescient.

The broader pattern? InfoFi shifted from a model that paid people to produce information (which incentivizes spam) to one that prices information quality (which incentivizes accuracy). As Benzinga framed it: the bet on InfoFi changed "from paying for posts to pricing credibility."

Why InfoFi Matters More Than Ever

Despite the January setback, several structural forces make InfoFi increasingly relevant.

The AI agent explosion demands trusted data. With 250,000+ daily active on-chain agents executing transactions worth billions, autonomous systems need machine-readable credibility signals. An AI agent deciding whether to allocate capital to a DeFi protocol needs more than price data — it needs reputation, audit history, and community sentiment, all programmatically verifiable. Intuition's attestation graph and Cookie DAO's agent index are early attempts to provide exactly this.

Information asymmetry is crypto's persistent failure mode. From the LUNA collapse to the FTX fraud, crypto's biggest disasters shared a common thread: critical information was either hidden, distorted, or impossible to verify. InfoFi protocols that can create transparent, staked information markets directly address this systemic vulnerability.

Prediction markets proved the concept. Polymarket's success — accurately calling elections, policy decisions, and market moves — demonstrated that financial incentives can extract genuine information from crowds. InfoFi extends this logic from binary yes/no questions to continuous reputation scoring, content quality assessment, and knowledge verification.

DWF Labs and institutional interest. DWF Labs positioned InfoFi as solving Web3's "information overload" problem, and institutional research desks increasingly recognize that whoever controls the information layer controls the alpha. As passive crypto data (price feeds, TVL metrics) becomes commoditized, the premium shifts to curated, contextualized intelligence.

The Road Ahead: Three Questions That Will Define InfoFi

The InfoFi sector faces existential questions that will determine whether it becomes a permanent infrastructure layer or a footnote in crypto history.

Can InfoFi survive without social media platforms? The X ban proved that building on another company's API is an existential risk. Protocols that own their own data infrastructure — like Intuition's L3 chain — have a structural advantage. Expect the next generation of InfoFi to prioritize platform independence.

Does tokenized attention create genuine price discovery, or just new forms of engagement farming? The early Yaps experiment showed how financial incentives can corrupt information quality. The pivot toward staking-based curation (where you lose money for backing bad information) is more promising, but it remains untested at scale.

Will AI agents become InfoFi's primary customers? If the AI agent economy reaches the $30 trillion scale that some analysts project by 2030, the demand for machine-readable credibility infrastructure would dwarf anything built for human users. InfoFi protocols that position themselves as the "trust API" for autonomous agents may find a massive, non-obvious market.

Conclusion

InfoFi's first year has been a compressed version of crypto's broader evolution: a visionary thesis, a speculative frenzy, a painful reckoning, and — potentially — a more durable rebuild. The sector's total market cap hovering around $350-500 million makes it a rounding error in the $2.5 trillion crypto market. But the underlying question InfoFi tries to answer — how do you create trustworthy information in a trustless world? — is arguably the most important unsolved problem in Web3.

Vitalik saw it in November 2024. The market is still figuring out whether it agrees.

BlockEden.xyz provides high-performance RPC and API infrastructure for the chains where InfoFi protocols are deploying — including Base, Arbitrum, and Ethereum. Explore our API marketplace to build on the infrastructure layer that powers the next generation of information finance.