# ASI Alliance: What the Merged Tokenomics Actually Mean > FET, AGIX, and CUDOS merged into one token. What each holder got, how the merged supply works, and why Ocean's lawsuit and exit changes the investment case. - Published: 2026-03-10 ยท Updated: 2026-10-07 - Canonical: https://ownyourmind.ai/tokenomics/asi-alliance-merged-tokenomics/ - Cite as: Own Your Mind (ownyourmind.ai), licence CC BY 4.0. Link the canonical URL. Charts and interactive widgets are omitted here; project data is at https://ownyourmind.ai/api/scores.json. --- ## What happened In March 2024, Fetch.ai, SingularityNET and Ocean Protocol announced they would merge into the Artificial Superintelligence Alliance. The pitch: combine three complementary AI projects (agents from Fetch, AI marketplace from SingularityNET, and data from Ocean) under one token and one governance structure. CUDOS joined in October 2024, contributing 30 validators. Then in October 2025, Ocean Protocol withdrew, alleging governance failures and strategic misalignment. Fetch Compute and three individuals filed a putative class action in New York the following month (*Fetch Compute, Inc. v. Pon*), alleging Ocean's Foundation converted 661 million OCEAN into 286 million FET and sold approximately 263 million FET (over 10% of circulating supply) without disclosure. That case is still open: the S.D.N.Y. docket (1:25-cv-09210) showed no termination when we checked on 24 September 2026, and media reports of a settlement aren't borne out by it. The reputational damage is done either way. The alliance continues as Fetch.ai + SingularityNET + CUDOS, but it lost its strongest data sovereignty component when Ocean walked. I don't hold FET. The governance breakdown that caused a founding member to leave within 18 months raises questions I haven't been able to answer from public information. For individual project assessments, see our [Fetch/ASI review](https://ownyourmind.ai/projects/fetch-asi/) and [Ocean review](https://ownyourmind.ai/projects/ocean/). For how FET compares to other DeAI tokens on value capture, see [MOR vs TAO vs FET](https://ownyourmind.ai/tokenomics/mor-vs-tao-vs-fet/). ## The conversion ratios All tokens consolidated under the **FET ticker** (not renamed to ASI on exchanges, despite the alliance name): Migration isn't 100% complete. OCEAN and AGIX still trade independently: OCEAN has roughly 241.5M tokens circulating on CoinGecko's count, AGIX approximately 250 million. No public deadline for conversion has been documented. Given Ocean's withdrawal, the status of remaining OCEAN-to-FET conversion is practically uncertain. (figures as of 2026-10-19) ## The merged supply The distribution of merged supply: The AGIX and OCEAN allocations (56.17% combined) represent community token swaps, not new insider supply. But approximately 600 million FET was minted to absorb OCEAN holders, which diluted original FET holders. Vesting runs to **2050**: a 24-year schedule for founders that's longer than most projects survive. ## What FET actually does FET has genuine multi-layer utility across the merged network: **Staking.** Secures Fetchhub-4 (Cosmos SDK chain with CometBFT consensus), up to 91 validators, with 74 bonded on 7 October 2026. Current APY approximately 5.5-8%. 21-day unbonding period. Third-party staking sites estimate 480-560 million FET is staked (21-25% of circulating supply). Standard Cosmos slashing: 5% for double-signing, 0.01% for downtime. **Agent services.** Pay for agent registration on the Almanac contract (renewed every 48 hours) and agent-to-agent transactions on Fetch.ai's Agentverse platform. **Compute.** GPU inference payments on ASI Cloud, which launched September 2025 and exited beta December 2025. Claims approximately 50% cheaper than AWS. I haven't verified this independently. **AI marketplace.** Payment for AI services on SingularityNET's marketplace (live since 2017, Ethereum and Cardano). **Governance.** Vote on alliance proposals including merger approvals, new member additions and tokenomics changes. **Burn mechanism.** The Earn and Burn programme is meant to use ecosystem fees to buy and burn FET. On chain there have been two burns, both from one Safe: 5 million FET on 10 January 2025 (0.18% of supply) and 109,350 FET on 5 November 2025, none since and none traceable to product fees. A $50 million buyback announced in June 2025 has no execution evidence. ## The products The alliance operates several platforms that are still largely separate: **Fetch.ai Agentverse.** Agent development and deployment platform with no-code browser-based templates. Agents register on-chain and can interact autonomously. It's the most mature product in the alliance. **SingularityNET marketplace.** AI services marketplace live since 2017. Developers publish and consume AI models. Usage volumes aren't publicly disclosed, which is a concern for a platform that's been live for nine years. **ASI Cloud.** GPU rental and AI model access. Permissionless. Includes ASI-1 Mini, described as the "first Web3-native LLM" with Knowledge Graph integration. **ASI:Create.** Closed alpha since February 2026. A launchpad for building, funding and deploying AI agents and applications. **ASI:Chain.** The convergence layer: a blockDAG L1 using the MeTTa programming language and Atomspace knowledge graph. DevNet launched November 2025. **No mainnet timeline published.** It's intended to unify all the above under one protocol, but it's pre-production. ## What Ocean's exit means Ocean Protocol was the data layer of the alliance. Its compute-to-data capability (running AI models on data without revealing the underlying dataset) was the strongest data sovereignty feature. With Ocean's withdrawal, that capability is no longer part of the alliance. The remaining components are agents (Fetch), AI marketplace (SingularityNET) and validators (CUDOS). These overlap more than they complement. The thesis that three distinct capabilities would combine into something greater depends on all three being present. Two and a half is a weaker proposition. The lawsuit is still live, and none of its allegations has been adjudicated. What it alleges is not reassuring: a founding member converting and selling over 10% of circulating supply without disclosure, surfacing only through litigation, would expose governance failures in how the merger was structured. Ocean's counter-allegations about governance not matching the "original decentralised vision" point to structural problems in the alliance itself. For a broader discussion of [what counts as decentralised AI](https://ownyourmind.ai/why-deai/what-counts-as-decentralised-ai/), the governance standards are worth examining. Whatever the court decides, a founding member leaving within 18 months isn't a sign of health. ## Next-generation tokenomics A whitepaper from superintelligence.io outlines a two-tier token model: - **ASI**: the base token (current FET) - **stakedASI**: a staked derivative with additional utility Shard-based economics with fee distribution: a portion burned, a portion to validators and treasury, a portion to staking rewards. Simulations show circulating supply declining by over 6 million ASI over 36 months with a target of approximately 11% annual staking yield funded by fees and buybacks rather than new issuance. Health-driven automated controls would increase burn rates during stress and hibernate emissions when necessary. Well-designed on paper. No implementation timeline has been published. ASI:Chain, which would host this model, has no announced mainnet date. ## What I'm watching **1. ASI:Chain mainnet.** The convergence layer that is supposed to unify all products under one protocol. Until it launches, the alliance is three separate products sharing a token. No timeline published. **2. The Ocean litigation.** The lawsuit is unresolved, and its allegation that 263 million FET was sold without disclosure hasn't been tested in court. The episode sits behind part of the price decline of more than 90%, and the reputational damage to the alliance's governance credibility persists. **3. Usage metrics.** None of the platforms (Agentverse, SingularityNET marketplace, ASI Cloud) publish adoption or revenue data. For a combined network that's been live in various forms since 2017, this opacity is a red flag. **4. The burn programme.** The first burn was 5 million FET (0.18% of supply) and the second, ten months later, 109,350 FET. Meanwhile Fetchhub-4 mints new FET at 3% a year to stakers. Burns would have to become regular and fee-funded to change the supply picture; at the current pace, they're symbolic. **5. Governance reform.** The governance structure that caused one member to leave hasn't been visibly changed. Until the alliance demonstrates it can make decisions transparently and retain members, the governance risk remains. ## The honest assessment The ASI Alliance has genuine technical talent. Ben Goertzel is one of the most credible AGI researchers alive. Fetch.ai's agent architecture is functional and differentiated. Combined patent portfolio and research output is substantial. But the execution has been troubled. A price decline of more than 90% from the merger announcement. A founding member's acrimonious exit. An unresolved lawsuit alleging undisclosed selling of over 10% of circulating supply. No published usage metrics across any platform. A convergence layer with no mainnet date. The merger thesis was that combining three specialised projects would create something greater. Losing the data layer and the associated lawsuit have damaged that thesis. What remains is a well-resourced agent platform (Fetch), a long-running but opaque AI marketplace (SingularityNET), and a validator set (CUDOS), sharing a token that's lost more than 90% of its value. The next-generation tokenomics design is genuinely interesting: shard economics, automated health controls, fee-funded yield. If ASI:Chain delivers and the alliance stabilises, the token economics could be compelling at its current depressed valuation (down more than 90% from ATH). But those are significant ifs, and I'm not comfortable betting on them until I see mainnet, usage data and governance reform.