ItsAI
SN32Detects AI-generated text, recognised as one of the most accurate AI detectors available
ItsAI takes AI text detection seriously enough to benchmark it hard, ship a public product, and keep the subnet loop focused on improving classification quality.
// Text AI detection, benchmarked hard.
ItsAI tackles a specific problem inside the Bittensor ecosystem: As generated text floods classrooms, publishing, and the web, weak AI detection tools create both false confidence and false accusations. Official sources describe it as a subnet where miners submit models and approaches that classify whether text is human-written or AI-generated, while validators benchmark outputs, compare performance across datasets, and score miners through the subnet incentive mechanism.
The simple version: It is like a decentralized GPTZero that keeps improving through competition.
Centralized equivalent: Think GPTZero or Originality.ai, but powered by Bittensor miners and validators.
How it works:
- Miners do submit models and approaches that classify whether text is human-written or AI-generated
- Validators check benchmark outputs, compare performance across datasets, and score miners through the subnet incentive mechanism
- The problem it solves: As generated text floods classrooms, publishing, and the web, weak AI detection tools create both false confidence and false accusations.
- The opportunity: High-quality detection can support educators, editors, platforms, and any workflow that needs provenance signals for text.
- The Bittensor advantage: Bittensor gives the subnet a live incentive loop, so detector quality can keep improving against new model behavior rather than freezing into one product version.
- Traction signals: ItsAI already operates a public detector site and reports benchmark results in official materials. The token trades near 0.00392, market cap is about 18,112 TAO, pool depth is roughly 7,775 TAO, and GitHub shows 550 commits from 8 contributors, though the last commit is older than some peers.
Category: Deepfake Detection and Security | Centralized Competitor: GPTZero, Originality.ai, Copyleaks
AI detection is messy. Users want confidence, but the underlying models keep changing. ItsAI has leaned into that reality with benchmark-heavy positioning and a public-facing product surface instead of hiding behind generic claims.
Mechanism:
The official README says the subnet focuses on detecting whether text is human-written or AI-generated, including sentence-level deep scan output. The same repo links mining, validating, and incentive docs, which suggests the scoring loop is an explicit part of the product rather than an afterthought. Safe wording here is simple: miners compete on detection quality, validators benchmark and score.
TAO.app and Supabase snapshot data show current pricing and demand conditions, but the mechanism and product-status claims above were kept anchored to official repos, docs, and websites. On the market side, ItsAI trades around 0.00392 TAO with roughly 18,112 TAO in market cap and about 7,775 TAO in pool depth. Seven day net flow sits near 625 TAO, which suggests recent demand has been constructive.
- Development Stagnation: The latest GitHub snapshot is less fresh than the fastest-moving subnets in this cohort. Product traction helps, but code velocity still matters.
- Competition: AI detection is brutally competitive, and benchmark leadership can disappear fast when foundation models shift.
- Execution: False positives are a reputational landmine. Detection products only win if users trust them in real decision settings.
