Yes — the claim by @kripto_kas1 is technically accurate on the base-layer blockchain, but it’s only half the story and misses critical context. Here’s the no-assumptions, fact-checked breakdown: Is Bittensor on Proof of Authority (PoA)? Yes — the root blockchain (Subtensor) currently runs a PoA-style consensus mechanism. •Subtensor is built on the Substrate framework (Polkadot tech stack). •Block production uses Aura (round-robin authoring by a pre-defined set of “authorities”) + GRANDPA for finality. •The authority set (the nodes allowed to produce and finalize blocks) is currently operated/controlled by the Opentensor Foundation. This is explicitly documented as Proof of Authority. •Official docs confirmation (as of April 11, 2026): “In Bittensor, the work of validating the blockchain is performed by the Opentensor Foundation on a Proof-of-Authority model.”6 This is not hidden — it’s openly stated in FAQs, multiple third-party analyses, and regulatory filings.589 Important nuance: Two separate consensus layers Bittensor has two distinct consensus systems — this is where most criticism (including this post) gets misleading: Layer Mechanism What it does Centralized? Base blockchain (Subtensor) PoA (Aura + GRANDPA) Produces blocks, records stake, weights, emissions Yes (currently) Intelligence layer (subnets) Yuma Consensus (YC) Scores miners/validators, distributes TAO emissions based on performance No — fully stake-weighted & decentralized •Yuma Consensus (on-chain, described in docs and evolved from the original whitepaper’s peer-ranking system) is the novel part: validators stake TAO, submit weight matrices ranking miner performance, and YC algorithm computes emissions. It assumes honest majority stake and is resistant to collusion. The whitepaper focuses almost entirely on this incentive/ranking layer (peer-ranking via neural nets, trust matrix, consensus term C, etc.). It never mentions PoA because the whitepaper was high-level architecture, not implementation details of the chain.016 •The PoA base layer is a temporary progressive-decentralization choice for security and stability in the early years (common pattern — see early Ethereum testnets, many L1 launches). Current status & roadmap (as of April 11, 2026) •The network is still on PoA. •Transition to Nominated Proof of Stake (NPoS) has been publicly planned and is described as “coming this spring” in documentation and community discussions (some sources from early 2026). No confirmed mainnet flip yet today, but governance upgrades are actively being discussed amid the Covenant drama.24 This PoA phase is exactly why critics (and some KOLs) call it “semi-centralized” — the Foundation can theoretically censor or pause the chain. However, no such event has ever occurred, and the intelligence production (128 subnets, real revenue, Yuma-driven emissions) runs permissionlessly on top. Why this question is surfacing now The Covenant AI subnet founder drama (one founder dumping bags and exiting) triggered a wave of “is Bittensor actually decentralized?” posts. @kripto_kas1’s question is part of that FUD cycle. It’s a fair technical point, but ignores that: •One subnet out of 128 had a founder issue → the other 127 kept producing. •The protocol is already shipping fixes (lock-based ownership, Senate governance upgrades). •This is classic “construction site” phase the whitepaper and founders always described. Bottom line: @kripto_kas1 is correct on the technical detail (Subtensor = PoA today), but calling the entire Bittensor network “semi-centralized” is an oversimplification. The value proposition — decentralized intelligence pricing via Yuma Consensus and the TAO incentive flywheel — is fully live and permissionless. This is progressive decentralization in action, not a scam or broken thesis.