X Personally Terminates InfoFi Incentive Model, Marking the End of the "Mouth Farming" Era

Original | Odaily (@OdailyChina)

Author | Ethan (@ethanzhang_web3)

Last night, a product update from X's official account caused significant ripples within the InfoFi community.

At 22:39 on January 15, the X platform announced the revocation of API access for InfoFi applications, immediately impacting multiple apps reliant on "posting incentives." With the API cut off, some projects announced the suspension of related features or adjustments to their business direction. Tokens associated with InfoFi saw significant declines, with several InfoFi-related tokens (KAITO, COOKIE) recording double-digit percentage drops within a short period. Community members offered a rather blunt summary of the situation—"The era of mouth-farming is over."

The severe reaction from InfoFi applications and their tokens indicates that this change's impact far exceeds a routine rule adjustment. It has altered the operational foundation for related applications and triggered a chain reaction in the market. This is not a minor tweak but a clear statement from X regarding a specific type of application model.

What Happened: X Formally Rejects the InfoFi Incentive Model

This time, X did not leave much room for interpretation for InfoFi.

X's Product Lead, Nikita Bier, posted on the platform stating that X is revising its Developer API policy and will no longer allow any applications that "reward users for posting on X" to continue accessing the API. In his statement, such applications were directly named as "infofi" and identified as one of the primary sources of AI-generated spam and reply pollution on the platform recently.

Unlike the platform's previous governance approach of "announce first, observe later," X's action this time was quite direct—API access for the relevant InfoFi applications has already been revoked. The official reason given is not complicated: external incentive mechanisms are driving a flood of task-based, templated content into the information feed, severely impacting the platform experience. X believes that once bots realize "there's no money to be made from posting anymore," the content environment will quickly self-correct.

It is worth noting that Nikita Bier specifically added a weighty statement: InfoFi applications had previously paid millions of dollars for API access, but X does not need this revenue.

This statement itself almost serves as a qualitative judgment on InfoFi's business model. Judging by the enforcement strength and official wording, this adjustment is not targeting individual projects abusing the API. Instead, it is X's clear and unavoidable rejection of the core InfoFi model where "external incentives directly intervene in platform content production."

The "transition plan" offered by X's official team for developers whose accounts were terminated is equally thought-provoking: the platform will assist them in transitioning their business to Threads and Bluesky. In other words, X did not attempt to reform or absorb this incentive mechanism but explicitly chose to move it entirely out of its own ecosystem.

What's Being Rejected is Not Content, But InfoFi's Incentive Pathway

If viewed solely from the official statement, this adjustment might seem like routine governance targeting AI-generated spam. However, within the InfoFi context, this reason is clearly insufficient to explain X's resolute stance.

The key issue may not be "whether the content has value," but rather who produces the content and for what reason. The core logic of InfoFi is to directly drive users to complete actions like posting, replying, and interacting on the platform through external token or point incentives. While this model indeed boosts short-term activity, it also rapidly transforms content production into "task execution." Posting is no longer about expressing opinions but a necessary step to claim rewards.

When incentives themselves operate outside the platform's governance system, the platform inevitably loses control over the motivation and quality of content. InfoFi applications do not care whether a reply adds informational value; they only care if it meets the "settlement" criteria. For X, this means the information feed is being taken over by an external economic system.

From this perspective, AI spam is a consequence, not the cause. What truly touches X's bottom line is the structural problem of "a third-party incentive layer being directly embedded into the platform's content distribution system." Once this model is tacitly permitted, the platform's content order, recommendation logic, and even user relationships would gradually be influenced by incentive designers.

This also explains why X left almost no room for InfoFi to adapt in this adjustment. It signifies that, in X's judgment, InfoFi is not an ecosystem participant that needs correction but rather a content production pathway that is no longer permitted to exist.

Thus, this API purge is X's proactive reclamation of its content sovereignty: when external incentives conflict with the platform experience, X chose to cut off the former rather than cede control over the information feed.

From "Shutdown" to "Restructuring": The Collective Pivot of InfoFi Projects

X's API purge did not remain at the policy level; it quickly triggered a chain reaction among InfoFi projects.

According to Odaily, the first to provide a clear response was Cookie DAO. After communicating with the X team regarding API and usage policies, the team announced the official shutdown of the Snaps platform and the termination of all ongoing creator incentive campaigns. Cookie stated bluntly in the announcement that this was a "difficult and sudden" decision, but the motivation was not to abandon InfoFi but to ensure its data layer and core products remain compliant.

Judging from the statement, the shutdown of Snaps appears more like a passive, loss-cutting choice under the impact of the event. On one hand, Cookie emphasized that it always uses official data sources and remains an enterprise-level API customer of X. On the other hand, the team also clearly stated that InfoFi is undergoing structural changes, and whether Snaps can exist in a "new form" still depends on further guidance from X. This wording itself reveals high uncertainty about the sustainability of the original incentive model.

In contrast, Kaito's adjustment appears more proactive. Kaito announced the shutdown of Yaps and its incentive leaderboard while launching the new Kaito Studio, explicitly bidding farewell to the "open, permissionless incentive distribution" path. According to the official statement, Kaito Studio will more closely resemble a traditional tiered marketing platform, where brands select creators for collaboration based on established criteria, and the covered platforms will expand from X to multiple social channels like YouTube and TikTok.

In explaining the reason for the pivot, Kaito did not shy away from the inherent problems of the InfoFi model. It pointed out that even after continuously raising thresholds and introducing screening mechanisms, low-quality content and engagement farming remained difficult to avoid. After communicating with X, the team also agreed that a "completely permissionless incentive distribution system" no longer aligns with the common needs of the platform, brands, and creators. Reading between the lines, the termination of Yaps might be an active abandonment of the original InfoFi path.

Regardless, viewing these two events together reveals a clear trend: when the platform layer explicitly tightens interface and incentive boundaries, InfoFi projects must either pause aggressive tactics and revert to data and tool attributes or completely restructure their business logic, moving closer to models resembling traditional marketing and content collaboration.

For now, although token prices have fluctuated, a "collective collapse" of InfoFi projects has not occurred. What is certain is that the playbook relying on platform APIs and using external incentives to directly drive posting and interaction is becoming increasingly difficult to sustain.

Conclusion: The Mouth-Farming Era Ends, But InfoFi's Questions Remain

Judging from the reactions of InfoFi projects, this change is not simply a "ban" or a "failure." Whether it's Cookie returning to its data layer positioning or Kaito pivoting towards a Studio model closer to traditional marketing, it indicates that: InfoFi has not disappeared; it just can no longer exist in the form of "in-platform incentive arbitrage."

The so-called "end of the mouth-farming era" does not mean the end of content being quantified or influence being priced. What has ended is that open incentive pathway reliant on APIs, where the act of posting and replying itself serves as the settlement object. Against the backdrop of platform sovereignty tightening again, the marginal space for this model is rapidly shrinking.

As for migrating to Threads or Bluesky, it seems more like a buffer solution than an answer itself. The real question is whether future InfoFi can find its own irreplaceable value position without taking over the platform's content production rights.

X is merely the first platform to explicitly press the button, but the signal it has sent is clear enough: Content sovereignty is returning to the hands of the platforms.