Islamic Finance Principles Assessment
Riba — Does affine involve interest?
Affine's core design shows no evidence of interest-bearing mechanics: it is a performance-based ML incentive contest rather than a lending or yield product. Rewards flow from network emissions tied to model performance, not fixed interest on deposits. For Muslim investors, the absence of any disclosed riba structure is a positive, though the lack of detailed treasury disclosure leaves some ambiguity.
Assessment: Moderate Riba
Score: 62.5/100
Our methodology examines 10 criteria to evaluate how well affine avoids interest-based mechanisms.
No source describes a distinct protocol-level revenue stream for SN120 beyond its emission schedule: roughly 2.31% of supply to miners, 2.31% to validators/stakers, 1.01% to the subnet owner, and a 2.86% AMM reserve injection, with genesis around June 26, 2025. There is no mention of a treasury holding interest-bearing instruments, no lending desk, and no debt-based income stream described anywhere in the available material. The protocol appears structurally to be a pure computation/ranking mechanism rather than a financial intermediary, which limits direct riba exposure, though full treasury composition remains undisclosed.
Rewards for SN120 are explicitly performance-driven rather than fixed: a challenger model must outperform the reigning champion across every evaluation environment to unseat it, with a fresh task pool sampled roughly every 24 hours. This winner-takes-all structure ties payouts to genuine computational output rather than to a predetermined interest rate on staked capital. "Validator & Staker Emissions" exist as part of the allocation, but the underlying mechanic — variable reward tied to contribution and outcome rather than a guaranteed return — is structurally distinct from riba-based lending arrangements.
Gharar — How much uncertainty does affine involve?
Uncertainty here is significant, driven less by the contest mechanism itself and more by disclosure gaps and identity confusion in the market. Open-source code and a clearly stated evaluation methodology reduce some ambiguity, but the absence of named leadership, audits, and full tokenomics details increases it substantially. On balance, this is a project where informational uncertainty, not the mechanism's own design, is the dominant concern.
Assessment: Excessive Gharar (High Uncertainty)
Score: 46.8/100
Our methodology examines 15 criteria including team transparency, audit quality, and governance.
The research surfaced four entirely distinct projects using the "Affine" name, and for SN120 specifically no accountable individuals or leadership team could be identified beyond the "AffineFoundation" GitHub organization. The code itself is publicly hosted and the contest logic (sybil-proof, decoy-proof, copy-proof, overfitting-proof champion mechanism) is described in reasonable technical detail, which is a genuine transparency positive. However, the absence of named, verifiable individuals behind the subnet, combined with pervasive brand confusion across unrelated "Affine" entities, makes independent due diligence unusually difficult for investors.
No audit of SN120's subnet mechanics or on-chain contracts could be confirmed in the available sources. Audits located under the "Affine" name — Certik and Quantstamp reviews of "Affine Restaking," and a Halborn review of "Multiplyr/Affine DeFi" — belong to an unrelated liquid-restaking startup, not this Bittensor subnet. This is a real and specifically named gap: an unaudited protocol handling real value distribution carries elevated gharar, and the unresolved vesting-cliff percentage from the October 2025 unlock only compounds the uncertainty around terms.
Maysir — Does affine involve gambling or speculation?
Affine does not resemble a gambling product in its own design; it is structured as a competitive machine-learning benchmark with productive output. Some uncertainty around rewards exists, but this reflects competitive performance risk rather than a wager on chance. The final take is that the base mechanism is not maysir, though speculative trading of the token in secondary markets is a separate, third-party matter.
Assessment: Moderate Maysir (High Risk)
Score: 56.8/100
Our methodology examines 11 criteria to determine whether affine is a gambling instrument or a genuine economic tool.
SN120's actual function is a winner-takes-all model-improvement contest: miners submit HuggingFace models that are evaluated against a reigning champion across coding and reasoning tasks, and only strict, comprehensive outperformance dethrones the champion. This produces a tangible output — improved AI models — rather than a payout determined by random chance. Emissions reward genuine computational contribution and measurable skill, distinguishing the mechanism from a betting pool even though outcomes are uncertain in advance, as they would be in any competitive R&D process.
Weighed against this genuine utility, the sources provide no market data, price history, or trading volume specific to the SN120 token, making it impossible to assess actual secondary-market speculative behavior from the available research. As with any freely traded token, speculative trading by third parties may occur, but this potential misuse does not stem from the protocol's own design and should not be read as evidence against the underlying contest mechanism. The bigger practical concern remains verifying which "Affine" a given market listing actually represents.
The Full 27-Point Screening
1. Legitimacy (4 criteria)
| Criterion | Score | Analysis |
|---|
| Team Transparency | 25/100 | Sources show only a GitHub organisation name (AffineFoundation) for SN120, with no named, credentialed individuals directly tied to the subnet identified, amid confusion with several similarly-named but unrelated "Affine" ventures. |
| Fraud & Scam Risk | 45/100 | No fraud, hack, or rug-pull reports specific to SN120 were found, but the pervasive naming confusion across multiple unrelated "Affine" projects makes independent verification of trust signals difficult. |
| Use Case Legitimacy | 75/100 | The protocol has a clearly documented technical use case: an incentivized reinforcement-learning contest that rewards miners for genuine model improvements. |
| Ethical Practices | 90/100 | The subnet's own design is a machine-learning benchmarking/incentive mechanism with no inherent connection to a prohibited industry. |
Summary: The sources show a genuine technical AI-incentive project on Bittensor but no named, accountable team for the SN120 subnet specifically, amid confusing overlap with several unrelated "Affine" branded ventures.
2. Project Operations (9 criteria)
| Criterion | Score | Analysis |
|---|
| Core Protocol Business | 90/100 | The base protocol operates in AI/machine-learning model evaluation, a sector with no Shariah prohibition indicated by the sources. |
| Transaction Fees | 50/100 (low evidence) | The sources do not describe how transaction fees on SN120 are handled (burned, retained, or distributed), so this could not be established. |
| Treasury Assets | 50/100 (low evidence) | No information on treasury composition or holdings for SN120 appears in the sources. |
| Revenue Model | 60/100 | The token's economics appear emission-based rather than interest-based, but no explicit revenue model is described, so this is inferred rather than confirmed. |
| Transparency | 80/100 | The subnet's code is publicly available on GitHub under the AffineFoundation organisation, supporting an open-source, transparent development model. |
| Governance | 35/100 | The presence of dedicated "Subnet Owner Emissions" suggests some centralisation around a subnet owner, but no detailed governance structure is described in the sources. |
| Launch Fairness | 45/100 | A genesis date and emission-based allocation are given, but details needed to judge launch fairness (e.g., insider pre-allocations) are incomplete, including an unresolved cliff percentage. |
| Token Distribution | 40/100 | Percentage allocations to miners, validators, subnet owner and an AMM reserve are listed, but the figures are inconsistent/unclear in the source, limiting confidence in how broadly the token is actually distributed. |
| Speculation/Utility Ratio | 60/100 | The protocol has demonstrable technical utility (ML incentive contest), but the sources provide no data on actual trading/speculative activity versus utility usage for the token. |
Summary: SN120 is an open-source, incentivized machine-learning contest between miner-submitted models, with emission-based allocations to miners, validators/stakers and the subnet owner, though fee handling, treasury and governance details are largely undocumented.
3. Financial Health (4 criteria)
| Criterion | Score | Analysis |
|---|
| Protocol Revenue | 60/100 | No interest-based revenue mechanism is indicated; income to participants appears to come from token emissions rather than lending, though this is inferred rather than explicitly confirmed. |
| Financial Status | 45/100 (low evidence) | No market capitalisation, price stability, or financial health data specific to SN120 could be found in the sources. |
| Interest Assessment | 90/100 | The base protocol is described purely as an ML-evaluation/incentive network with no lending or borrowing function at the protocol level. |
| Audit Quality | 15/100 | No audit report for the SN120 subnet itself was found; audits identified in the sources belong to differently-named, unrelated "Affine" DeFi entities, confirming the absence of a verifiable audit for this specific token. |
Summary: No protocol revenue model, financial stability data, or audit specific to SN120 could be found, and the base protocol does not appear to offer native lending or interest-based functions.
4. Token Economics (5 criteria)
| Criterion | Score | Analysis |
|---|
| Token Purpose | 80/100 | The token functions as a utility/incentive instrument rewarding genuine model-improvement work rather than serving as a purely speculative meme asset. |
| Governance Rights | N/A | No holder governance rights are described for the token, and the sources give no indication this absence is itself problematic for a mining-incentive subnet token. |
| Rewards Distribution | 85/100 | Rewards are explicitly variable and performance-based, tied to whether a miner's model can dethrone the reigning champion across evaluation environments. |
| Speculation Controls | 30/100 | Beyond a partially-documented vesting cliff, no meaningful anti-speculation mechanisms (e.g., burn, buy-back, holding limits) are described in the sources. |
| Asset Backing | 55/100 | The token appears to derive value from genuine network utility (ML task incentives) rather than a hard asset, but this is inferred rather than explicitly stated. |
Summary: The token rewards genuine model-improvement performance in a variable, winner-takes-all structure rather than functioning as a meme or governance token, though anti-speculation design and asset backing are thinly documented.
5. Staking Mechanism (5 criteria)
| Criterion | Score | Analysis |
|---|
| Mechanism Type | 30/100 (low evidence) | While "Validator & Staker Emissions" indicate some staking/delegation exists, the sources give no detail on custody, delegation type, or lock-up terms. |
| Islamic Contract Classification | 30/100 (low evidence) | No source classifies the staking arrangement under any Islamic contract framework, leaving this unresolved. |
| Rewards Structure | 55/100 | Staking rewards appear to derive from network emissions tied to subnet activity rather than a fixed guaranteed rate, but the exact reward source and formula are not detailed. |
| Documentation | 20/100 (low evidence) | No documentation of staking terms, risks, or procedures for SN120 could be found in the sources. |
| Shariah Alignment | 30/100 (low evidence) | With mechanism, contract classification, and documentation all unestablished, a clear Shariah alignment judgment cannot be made from these sources. |
Summary: A validator/staker emission mechanism appears to exist, but its custodial nature, lock-up terms, slashing conditions, and Islamic contract classification are not documented in the available sources.
Overall Assessment: Affine (SN120) presents as a legitimate machine-learning incentive protocol with plausible utility, but significant gaps in team transparency, audit evidence, governance clarity, and staking documentation prevent a fully confident Shariah assessment from these sources alone.
Written by ThanvirFounder, Ex Director S&P Global Energy
Reviewed by Imam Omar SiddiqiShariah Scholar