Bittensor TAO Governance Crisis Triggers 25% Price Drop Amid Centralization Allegations

Bittensor (TAO) has suffered a severe market correction after Covenant AI, a prominent subnet developer, publicly exited the network. The departure was precipitated by accusations that co-founder Jacob Steeves maintained centralized control over the protocol's governance and economic incentives. This internal conflict caused TAO's price to plummet by more than 20% within 24 hours, erasing nearly $900 million in market capitalization.

Covenant AI founder Sam Dare alleged that Steeves unilaterally suspended emissions to their subnets, effectively cutting off their revenue stream during a dispute. On-chain forensics suggest that 38 of the 41 network upgrades between 2023 and 2026 were deployed from infrastructure controlled by Steeves. This data point has intensified scrutiny regarding the network's claim to be a truly decentralized marketplace for artificial intelligence.

In response to the crisis, Steeves and other co-founders have proposed a new governance feature known as the Locked Stake mechanism. This update aims to align long-term incentives by requiring subnet owners to lock their token holdings for a defined period. The proposal is intended to demonstrate commitment and prevent similar sudden exits that destabilize the network.

The fallout has also cast doubt on institutional adoption efforts, specifically Grayscale's pending application to convert its Bittensor Trust into a spot ETF. The governance turmoil highlights the material risks for investors who rely on distributed decision-making authority in decentralized protocols.

How Did The Covenant AI Exit Impact Bittensor's Market Value?

The exit of Covenant AI, which operated critical subnets including Templar, Basilica, and Grail, triggered a massive sell-off of TAO tokens. Dare's team reportedly sold 37,000 TAO tokens worth approximately $10.2 million, exacerbating the downward price pressure. This liquidation event triggered over $10 million in long positions to be liquidated across various exchanges.

The immediate market reaction saw TAO lose about a quarter of its value in a single day, a volatility spike not seen since late 2024. The broader ecosystem of Bittensor subnets also declined, with the Templar subnet specifically dropping 63% following the announcement. This contraction in subnet value underscores the fragility of the network's economy when key builders depart.

Analysts note that the departure severs a major demand driver for the TAO token, as Covenant AI had previously trained a 72-billion-parameter model distributed across 70 independent nodes. While the technical achievement of this training remains a proof of concept for decentralized AI, the governance dispute undermines the long-term viability of the project. The loss of such a significant subnet operator raises questions about the network's ability to protect builders if infrastructure can be weaponized against them.

What Governance Reforms Are Proposed To Restore Trust?

Bittensor co-founder Const acknowledged the pain caused to holders and apologized for the lack of implementation of the Locked Stake mechanism. He admitted that failing to introduce this feature earlier was an error that could have prevented the recent value drop. The proposed mechanism combines stake size with lock duration to create a publicly legible measure of commitment.

Under the new proposal, subnet owners would be required to lock their holdings for a fixed period, creating a metric where time plus stake equals trust. This design aims to solve the Web3 problem of measuring team commitment beyond simple ownership, offering investors predictability against rug pulls. The proposal will be presented to the community for a vote at the next open call.

Steeves confirmed that the mining community and former Covenant team members are organizing to sustain operations on affected subnets, specifically subnets 3, 39, and 81. The goal is to revive these subnets while awaiting a formal governance decision on the new locking requirements. This move is intended to set a new benchmark for decentralized AI projects by aligning long-term incentives with open-source development.

Why Do Centralization Allegations Threaten Institutional Adoption?

The allegations of centralized control strike at the core of the investment thesis for decentralized networks, which rely on the premise that no single actor can alter rules unilaterally. Evidence suggests that a triumvirate structure allows Steeves to make unilateral changes without consensus, contradicting the network's decentralization claims. On-chain data indicated that other co-signers often acted within minutes without documented discussion on recent upgrades.

For institutional investors, particularly those considering Grayscale's pending TAO ETF filing, this governance concentration represents a material risk. The investment case for decentralized networks relies on the premise that no single actor can alter rules unilaterally. Evidence of concentrated control undermines this premise directly, potentially complicating the approval process for regulated financial products.

The crisis highlights that price stability in protocols reliant on key subnet operators is fragile when governance disputes arise. Future growth will depend on resolving these structural tensions and proving true decentralization to institutional partners. While the technical proof of decentralized training stands, the governance structure remains a point of contention that could impact long-term adoption and trust in the network.

The broader AI crypto sector faced volatility during this period, with Bittensor subnets among the worst performers. However, the network's ability to pivot and implement the Locked Stake mechanism will be critical in determining its future trajectory. The market is now watching closely to see if the community can effectively govern the protocol without relying on a single central figure.