Bittensor Locked Stake: The Conviction Mechanism Explained

Bittensor is preparing one of the most significant governance upgrades in its history. The proposal, known as BIT-0011, introduces a feature called Bittensor Locked Stake. It adds a time dimension to staking and creates a new framework for subnet ownership based on verifiable, long-term commitment.

Jacob Steeves, known as Const, outlined the proposal during a live community session on the official Bittensor Discord on Thursday, April 16, 2026. The session followed the recent departure of Covenant AI and its founder Sam Dare from the ecosystem, an event that triggered a 25% drop in TAO price and raised urgent questions about subnet governance. The crisis arrived at a particularly sensitive moment for the network. Just days earlier, Grayscale had filed an amendment to its S-1 registration with the SEC, advancing its effort to convert the Bittensor Trust (GTAO) into a listed spot ETF. Grayscale also increased its TAO allocation within its AI fund to 43%.

What Is Bittensor Locked Stake?

The core idea behind Bittensor Locked Stake is straightforward. Participants can lock their alpha tokens on a subnet for a chosen period of time. This lock is entirely self-decided. Once locked, the tokens produce a conviction score that combines the locked amount with the remaining lock duration.

The conviction score starts at 100% of the locked value. As the lock period approaches expiry, the score decays linearly toward zero. Every 30 days, the network calculates a new conviction score using an Exponential Moving Average (EMA). This smoothing mechanism prevents anyone from briefly surging to the top with a large, short-term lock.

The participant with the highest conviction EMA on a given subnet becomes its owner. Locked tokens remain fully locked and cannot be unstaked as long as the conviction is active. This creates a publicly visible and cryptographically verifiable measure of commitment that replaces the need for human trust assumptions.

Why Bittensor Needs This Now

The timing of this proposal is directly tied to the Covenant AI crisis. When Sam Dare exited the Bittensor ecosystem on April 10, 2026, the sudden departure exposed a critical vulnerability. Subnet owners could accumulate significant alpha positions, build community trust, and then exit without warning, draining liquidity and damaging investor confidence.

Jacob Steeves acknowledged this vulnerability during the Discord session. He took responsibility for the delayed implementation and offered direct apologies to affected holders. He framed the situation as proof that human-led governance carries inherent risks, and that the protocol needs mathematical, protocol-level guarantees instead of reliance on personal trust.

In a notable detail, Jacob Steeves revealed that the Bittensor locked stake mechanism was among the last features Sam Dare himself worked on before departing the Opentensor Foundation. Const described the failure to implement it sooner as his “real error”, suggesting the mechanism could have prevented the crisis entirely.

The proposal addresses the core problem at its root. Subnet owners who lock their tokens for extended periods make their commitment transparent and measurable. Any attempt to exit would require waiting for the lock to expire, giving the market sufficient time to react and reprice.

How Subnet Ownership Becomes Dynamic

One of the most transformative aspects of BIT-0011 is that it makes subnet ownership competitive and dynamic. The current model does not require subnet owners to demonstrate ongoing commitment. An owner key can remain with the original registrant regardless of their actual engagement or activity level. BIT-0011 changes that entirely.

Under the new mechanism, anyone can lock alpha tokens on a subnet and build conviction. If a challenger accumulates a higher conviction score than the current owner, they become the new owner after the next 30-day evaluation cycle.

This introduces a healthy competitive dynamic. Subnet teams that demonstrate genuine long-term alignment through large, extended locks will maintain ownership. Teams that reduce their commitment or allow their locks to expire will face legitimate challenges from more committed participants.

Jacob Steeves confirmed that the rollout will begin with a grace period for new subnets, providing immunity from ownership challenges during their early development phase. Initial testing will take place on mature subnets like SN3, SN39, and SN81, the same subnets previously operated by Covenant AI that are now being revived by community members and former contributors.

What This Means for TAO and Alpha Holders

For investors holding TAO or subnet alpha tokens, the Bittensor locked stake mechanism offers several concrete benefits.

Transparency becomes the default. Every subnet owner’s conviction score will be publicly visible on-chain. Holders will be able to verify exactly how much stake an owner has locked and for how long. This eliminates the guesswork involved in evaluating whether a team is genuinely committed or preparing an exit.

Rug pull prevention moves from social enforcement to protocol enforcement. Locked tokens cannot be moved until the lock expires. A subnet owner who locks for 12 months provides 12 months of guaranteed presence. If their conviction decays, the community has time to act, whether by challenging ownership or adjusting their own positions.

Capital formation improves because investors can make more informed decisions. Teams that lock significant amounts for long periods signal serious commitment. This measurable conviction metric makes it easier to evaluate subnets and allocate capital accordingly.

The Bigger Picture for Bittensor Governance

Jacob Steeves positioned the Bittensor locked stake mechanism as just the first step in a broader governance roadmap. The initial rollout focuses on the lock mechanism itself, without hierarchical governance layers. As the system matures, the plan includes stronger governance at the subnet level and eventually at the root and chain level.

The long-term vision is ambitious. Bittensor aims to solve governance problems that remain unsolved across Web3 as a whole. The conviction mechanism creates a foundation where decisions are driven by provable commitment rather than reputation, branding, or informal influence.

Jacob Steeves reinforced during the session that the Bittensor protocol has always evolved through stress tests. Every exploit and failure since the network’s earliest days has led to improvements in the chain, refinements in incentive design, and stronger mechanisms for building functional decentralized AI markets. The Covenant crisis, while painful, is following this same pattern.

What Comes Next

The BIT-0011 proposal is currently in community discussion. Jacob Steeves presented the formal design during the Discord session and will continue taking feedback from the community. The timeline for implementation has not been finalized, but the initial testing phase on mature subnets is expected to begin soon.

For the Bittensor ecosystem, this moment marks a turning point. The shift from trust-based ownership to conviction-based ownership has the potential to set a new standard for how decentralized networks handle governance and accountability. If implemented successfully, Bittensor locked stake could become a benchmark for the entire decentralized AI sector.

Frequently Asked Questions