Exclusive: A Peek into $TAO’s Conviction Framework

Exclusive: A Peek into $TAO’s Conviction Framework
Read Time:4 Minute, 46 Second

Bittensor’s most consequential governance proposal is no longer theoretical, and at the OpenDev call on May 5, 2026, the first two pull requests for Subnet Conviction were made public, giving the ecosystem its earliest concrete look at how the mechanism is actually being built. 

Doug from Taostats released two breakdowns the following day, one simplified for general users and the other (a little technical though!) for subnet owners. Together, they offer the cleanest picture currently available of what Conviction does, how it works, and why it represents one of the most structurally significant changes to Bittensor’s governance model since Dynamic $TAO. 

What Conviction Actually Is

The simplest framing of Conviction is that it locks subnet ‘$ALPHA’ tokens to a hotkey in a way that demonstrates long-term commitment to a subnet. 

Doug walked through the mechanics directly:

a. All Future Subnet Owner Emissions will be Locked into Conviction by Default: Owners receive 18% of every $ALPHA emitted per block, and going forward, that allocation flows directly into the Conviction pathway rather than into freely tradable supply

b. Locked $ALPHA Translates into Conviction Over Time, not Immediately: The maturity rate runs across roughly 90 days, with a 62-day half-life, meaning half of the locked $ALPHA shows up as Conviction after two months and approaches full conversion after roughly two years

c. The Hotkey Holding the Highest Conviction in a Subnet is Referred to as ‘The Subnet King:’ For now, the title is purely honorific, though there is active discussion about whether this could eventually translate into actual subnet ownership rights

d. New Subnets Registered After the Upgrade will See Their Initial $ALPHA Allocation Locked into Conviction Automatically: Existing subnets will not have any retroactive locking, though owners may voluntarily commit alpha to build conviction over time

The core point Doug emphasized for general users is that Conviction is not staking. While Staking generates yield and can be distributed across multiple validators, Conviction generates no yield, can only be committed to a single hotkey per subnet, and carries lock-up conditions that staking does not.

The Math Subnet Owners Need to Understand

For subnet operators specifically, the mechanics carry direct economic consequences worth modeling carefully:

a. Conviction Grows Continuously: With 0.18 $ALPHA emitted per block and locked into Conviction, owner Conviction compounds across every block. Over 200 days, an initial 200,000 $ALPHA lock can grow into approximately 330,000 $ALPHA of Conviction.

b. Initial Registration $ALPHA is Locked at the Start for New Subnets: Recent registrations have produced between 100,000 and 200,000 $ALPHA at the moment of subnet creation, and that initial pool now flows directly into Conviction rather than remaining liquid.

c. Existing Subnets Begin from Zero: Existing subnets without retroactive locking still cross 50,000 Conviction within roughly 100 days simply through the per-block emission flowing into the Conviction pathway.

d. Unlocking has Cascading Effects: When $ALPHA is unlocked, the locked balance and the Conviction value both drop by the unlocked amount immediately. Conviction begins regrowing on the new lower base, while the unlocked alpha decays into availability over a 30-day window

For owners running operational businesses, this means that paying salaries, funding development, or covering infrastructure costs now requires forward planning. 

A team needing accessible $ALPHA 30 days from now must initiate the unlock today and accept the corresponding drop in Conviction visible to the broader community.

Why the Average User Should Still Pay Attention

Conviction is primarily a subnet owner mechanism, and most general users will never lock their own $ALPHA into the system. 

That does not make the feature irrelevant for retail participants:

a. Conviction Levels Become a Powerful Trading Signal: Watching whether a subnet owner is steadily building Conviction or repeatedly unlocking it offers a direct read on operator commitment.

b. Public Unlocks Open Governance Conversations: When an owner reduces Conviction, the on-chain visibility forces an explanation, which gives the community real-time insight into how subnets are being run.

c. The Subnet King Concept may Evolve into Something More: Discussions are already underway about whether the highest-Conviction hotkey could eventually carry actual ownership rights, which would introduce a new dimension to subnet governance that capital-rich actors could exploit or counterbalance.

The broader effect, as Doug summarized it, is that Conviction shifts the conversation around subnet ownership from quiet trust to visible commitment. 

The mechanism does not guarantee good behavior, but it makes good behavior easier to verify and bad behavior much harder to execute silently.

The Mechanism Is Live in Code, Not Yet on Mainnet

Two important caveats sit underneath everything above. The first is that the code is genuinely early as the two pull requests reviewed at OpenDev represent the initial implementation, and Doug expects substantive changes before Conviction reaches mainnet. 

The second is that the rules around existing subnets remain undefined: Whether legacy subnet owners will be able to retroactively lock $ALPHA, whether any of their existing holdings will be subject to Conviction mechanics, and how the transition will be sequenced are all questions still being worked out.

What is already clear is that Conviction is the protocol’s structural answer to the Templar problem, and it represents a deliberate shift toward making subnet ownership a continuously earned position rather than a permanent title. 

Whether this is the final form of the mechanism or a first iteration on the way to something more refined, the conversation around subnet accountability is about to get significantly more measurable. For anyone watching how Bittensor governance actually evolves, the next few weeks of OpenDev calls will be the ones worth following.

Enjoyed this article? Join our newsletter

Get the latest TAO & Bittensor news straight to your inbox.

We respect your privacy. Unsubscribe anytime.

Be the first to comment

Leave a Reply

Your email address will not be published.


*