Covenant AI, the team behind the Covenant-72B model that drove TAO’s 90% March rally, has exited Bittensor — accusing founder Jacob Steeves of unilateral control over a network marketed as decentralized.

TAO, the native token of the Bittensor decentralized AI network, dropped over 15% on Thursday after Covenant AI — the team behind the network’s most high-profile project — announced it was leaving Bittensor over what it described as centralized control disguised as decentralized governance.

Covenant AI founder Sam Dare published a public statement accusing Bittensor founder Jacob Steeves (known on-chain as “Const”) of maintaining effective control over the network’s “triumvirate” governance structure, resisting any meaningful transfer of authority, and deploying changes unilaterally. Dare specifically alleged that Steeves suspended emissions to Covenant’s subnets — eliminating the team’s income — and orchestrated large TAO token sales as financial pressure during the dispute. “It is decentralization theatre,” Dare wrote.

Covenant AI built the Covenant-72B model, the largest decentralized LLM pre-training run in history, across more than 70 independent contributors on commodity hardware. That achievement had driven TAO’s 90% rally in March and earned public praise from Nvidia CEO Jensen Huang and investor Chamath Palihapitiya.

This story is an excerpt from the Unchained Daily newsletter.

On the same day as Dare’s announcement, a newly launched site called Tao Papers published internal documents and on-chain forensics from multiple whistleblowers. The site claims that of 41 Bittensor network upgrades between 2023 and 2026, 38 were proposed, first-signed, and deployed from infrastructure controlled by Steeves — with the other two multisig signers co-signing within minutes and without public discussion.

Steeves responded by saying Bittensor is on the verge of launching “headless” subnets that operate as true commodities— though that statement did not address the specific allegations. The episode revives a structural concern for the broader AI token market: TAO’s valuation was heavily tethered to one team’s narrative, and that team has now left.