Bittensor Proposes “Conviction” to End Subnet Rug Pulls

Steady attention without excessive speculation.

Co-founder Jacob Steeves, known as Const, introduced the Bittensor Conviction mechanism during a Discord AMA on April 16, 2026. He described it as one of the biggest updates since Dynamic TAO. The proposal is now formalized as BIT-0011 under the title “Locked Stake and Conviction.”

What Triggered the Proposal

On or around April 10, 2026, Covenant AI founder Sam Dare exited the project. Dare sold approximately 37,000 TAO, worth about $10 million at the time, and publicly accused Const of centralization.

As a result, TAO’s price dropped 20-25%. Liquidations followed, and the community expressed outrage over what many called a “rug pull.”

Const framed the incident as a necessary “stress test” for the network. Bittensor has a history of improving through exploits, so he positioned Conviction as the structural fix that comes next.

Loading chart...

How the Bittensor Conviction Mechanism Works

The core formula is simple: Conviction equals Stake multiplied by Time. Any participant can voluntarily lock alpha tokens on a chosen subnet for a self-selected duration. According to simplytao.ai, the conviction score starts at 100% of the locked stake value. It then decays linearly to zero over the lock period.

Tokens cannot be unstaked while the lock is active. In practice, this creates a verifiable, on-chain commitment that replaces informal trust.

Every 30 days, the network computes an Exponential Moving Average (EMA) of each staker’s conviction score. This smoothing mechanism deters short-term manipulation. As a result, the staker with the highest EMA conviction becomes the subnet owner.

Ownership is therefore no longer permanent. When a founder stops performing, other stakers can collectively out-lock them. They can then take control of the subnet through pure on-chain competition.

Built-In Protections Against Abuse

The proposal includes several safeguards to prevent gaming. Specifically, new subnets receive an “immunity period” to bootstrap before ownership challenges can begin. This gives early teams time to build without facing hostile bids on day one.

The design also deters whale attacks economically. If a hostile actor locks large amounts of alpha to seize a subnet, the community can respond by selling unlocked tokens and migrating to a new slot. The attacker then holds locked, devalued tokens on a dead subnet. Meanwhile, the original team keeps its IP, code, and assets.

In addition, miners and investors can delegate their locked conviction to back trusted teams. This feature allows smaller holders to pool their commitment behind operators they support.

Transparency as a Market Signal

All locks, durations, and conviction scores live on-chain and remain fully public. Because of this, the market can see impending unlocks before they happen and reprice risk accordingly.

According to the Opentensor Foundation, this transparency eliminates the opacity that enabled the Covenant AI exit. Investors no longer need to guess whether a founder plans to stay committed. Instead, the data sits visible in real time for anyone to verify.

Teams can still raise capital by selling OTC futures on locked alpha. They receive cash now and deliver tokens later, so the lock does not prevent fundraising.

Rollout Plan and Current Status

The Bittensor Conviction mechanism is not yet live. It remains a working proposal under community discussion and refinement.

According to Const’s AMA, the rollout follows a phased approach. First, Conviction launches as an optional primitive on the chain. Then, initial testing targets compromised mature subnets 3, 39, and 81, where the community needs tools to reclaim control.

After that, the mechanism expands to other mature subnets with sufficient liquidity and community depth. New subnets stay protected during their immunity period. Const has not finalized a hard timeline for mainnet activation.

Community Response and Market Implications

Community sentiment on X and governance forums has been largely positive. For example, multiple explainer threads, visual breakdowns, and calls for rapid implementation appeared within hours of the AMA.

Several large TAO holders have publicly backed the proposal. They argue that Conviction strengthens decentralization and gives investors verifiable protection against founder exits. Some community members noted that the mechanism could restore confidence after the recent volatility.

The Opentensor Foundation published detailed mechanics and FAQs on X on April 17, 2026. The posts address concerns about hostile takeovers, delegation, and the immunity period for new subnets.

What Comes Next for Bittensor

The Conviction mechanism, if adopted, represents a shift in how decentralized networks handle governance. Instead of relying on legal agreements or social trust, on-chain commitment determines subnet ownership through locked capital and time.

BIT-0011 is still in draft form and may change based on community feedback. The design could move quickly given the urgency that the Covenant AI incident created.

This is not financial advice. TAO and subnet alpha tokens carry significant risk, and the Bittensor Conviction mechanism remains a proposal, not a deployed feature.