What the Covenant AI Exit Actually Revealed About Bittensor Governance

On April 10, Covenant AI founder Sam Dare announced his project was leaving Bittensor entirely, calling the network's governance "decentralization theatre" and accusing co-founder Jacob Steeves of exercising centralized control despite the protocol's decentralized branding. TAO dropped roughly 25-27% that day on $1.72B in trading volume — more than half the token's entire market cap changing hands in 24 hours.

The allegation and the response

Dare alleged that Steeves suspended emissions to Covenant's subnets, restricted its moderation rights, deprecated infrastructure it relied on, and applied pressure through token sales tied to their operational disagreement.

Steeves denied the core claims directly: "I do not have the ability to suspend emissions," he wrote, and said he sold some of his own alpha holdings on Covenant's three subnets because they weren't running — operating on what he described as near-100% burn code, meaning they were consuming emissions without doing real work.

Both sides are on the record, and the two accounts genuinely disagree on the underlying facts, not just the interpretation.

The structural point that doesn't depend on who's right

Whatever actually happened between Dare and Steeves, the dispute surfaced something worth sitting with regardless of which account you believe: Bittensor's protocol-level governance runs through a small multi-signature group — sometimes referred to as the "Triumvirate" — that holds real authority over parameters most token holders have no direct say in. That's a structural fact independent of this specific fight, and it's the actual risk to price into holding TAO or subnet alpha tokens: not that a dispute like this will happen again, but that governance is concentrated enough for one to matter this much when it does.

TAO has partially recovered since April. The governance structure that made a single founder dispute worth a 25% single-day move hasn't materially changed.