Fetch.ai (FET): Autonomous Economic Agents and the ASI Alliance

Opening paragraph

Fetch.ai started with an ambitious premise: what if software agents could autonomously negotiate, trade, and coordinate in markets without human intermediaries? Toby Simpson and Humayun Sheikh founded it in 2017 and launched mainnet in May 2019.

The core idea is that agents (AI-powered software) running on-chain can discover each other, negotiate terms, and execute transactions automatically. Instead of a person asking "who has data I need?" or "who wants to buy my goods?", the agent figures it out.

In October 2024, Fetch merged with SingularityNET and Ocean Protocol to form the Artificial Superintelligence (ASI) Alliance. The merged group is building an entirely new blockchain (ASI:Chain) designed specifically for this kind of agent work, using blockDAG consensus to handle high concurrency. That launches late 2026 if plans hold.

Today FET trades at $0.23 (April 2026), down from $1.40 in May 2024, reflecting delays, governance disputes, and investor skepticism about whether decentralized autonomous agents actually solve real problems.

History and founding

Fetch emerged from the conviction that blockchains weren't built for agent-to-agent coordination. Bitcoin and Ethereum work for humans pushing buttons. Fetch wanted infrastructure where agents could operate autonomously.

The platform launched via Binance Launchpad and built out marketplaces for agents to transact: data sales, supply chain negotiations, trading algorithms. By 2023 it hit $1B valuation, but adoption stayed niche—mostly enterprise pilots and theoretical use cases.

Then Ocean Protocol and SingularityNET joined forces with Fetch in late 2024. The thesis: three decentralized AI projects could combine resources and governance to build a blockDAG Layer 1 optimized for agent throughput. That consolidation hasn't gone smoothly.

Technical architecture

Current consensus: CometBFT (Cosmos SDK)

Fetch mainnet runs on Cosmos SDK using CometBFT (formerly Tendermint BFT). That gives 5-second block times, ~5-second finality, and about 50 TPS. Transactions can move between Cosmos chains via IBC, and smart contracts run in Rust via CosmWasm.

Planned ASI:Chain (late 2026)

The merger partners are building a custom blockDAG to replace sequential consensus. Traditional blockchains process transactions one block at a time. BlockDAG allows many transactions to settle concurrently.

The spec targets 100,000+ TPS and sub-second finality. The idea: when 100 agents are negotiating simultaneously, they shouldn't queue up and wait for block production. BlockDAG lets all of that happen in parallel.

Smart contracts and agents

Fetch supports two programming models. CosmWasm for traditional smart contracts (DeFi, tokens, governance). And the Autonomous Economic Agent (AEA) framework in Python for agents that live off-chain, register on-chain, and autonomously call contracts or transact.

ASI:Chain will add native primitives: agents can execute parallel transactions, use state channels for agent-to-agent payments, and attest off-chain computation with on-chain settlement.

Current bottlenecks

The 50 TPS limit means agents have to batch requests to save on fees, adding latency. State bloat is a long-term problem—old agent registrations and negotiation history accumulate. These limitations motivated the blockDAG redesign.

Ecosystem and adoption

FET supports data marketplaces (agents buying and selling datasets), supply chain agents (warehouse and logistics automation), DeFi agents (arbitrage bots, liquidation strategies), and some enterprise pilots (companies experimenting with automation). But most use cases are narrow (trading bots) rather than general-purpose. Real adoption remains limited.

The ASI Alliance also includes developer tools, wallet integrations, and exchange listings.

Exchanges, wallets, and infrastructure

FET trades on Coinbase, Kraken, OKX, and Binance with daily volumes around $200–$400M. Keplr is the native wallet (Cosmos), MetaMask bridges via Wormhole, Ledger works via the Cosmos app.

Tokenomics

FET has 2.3 billion in circulation out of a 2.7 billion cap. No halvings—inflation is predetermined and decreasing. The network is valued at $534M (rank 99).

Allocation was roughly 15% foundation/team (vested over 4 years), 25% community programs, 60% circulating. FET gets used for marketplace discovery, validator staking, governance voting, and agent payment settlement.

Governance and development

The foundation currently manages protocol upgrades and treasury, with community input on major decisions. The plan is to transition to a decentralized autonomous organization (DAO) when ASI:Chain launches.

But that transition hasn't gone smoothly. Ocean Protocol withdrew from the alliance in October 2025 over disputes about token allocation and treasury governance. Ocean accused Fetch of misrepresenting community token reserves; Fetch accused Ocean of converting and selling those tokens to inflate their own OCEAN supply. Legal disputes are ongoing. The blow-up eroded investor confidence in the merged alliance.

Core roadmap targets: ASI:Create platform beta (Q2–Q3 2026), ASI:Chain testnet launch (Q3 2026), agent payment system rollout (originally January 2026, likely delayed), and mainnet launch late 2026.

Regulatory status

The EU AI Act classifies autonomous decision-making systems as high-risk, requiring transparency, human oversight, and auditability. That's hard to guarantee in a decentralized agent network where no single entity controls all agents.

Most jurisdictions don't classify FET as a security. But if agents start managing customer funds (like robo-advisors), regulators may reassess that classification.

GDPR compliance is murky for agents that access personal data across multiple jurisdictions.

Controversies and risk factors

ASI Alliance fragmentation

Ocean's departure and the ongoing legal dispute damage the alliance's credibility. Public governance fights suggest structural problems and may scare off enterprise buyers.

Autonomous agent liability

Basic legal questions remain unanswered. If an agent makes a bad trade, who's at fault? The developer? The infrastructure provider? The end-user? Decentralized agents blur accountability. These ambiguities make institutions nervous.

Execution risk

BlockDAG consensus is technically novel but unproven at scale. IOTA and Hedera use it, but no production-grade blockDAG Layer 1 handles Bitcoin-level volumes yet. ASI:Chain is a bet on new technology.

Adoption risk

Real-world agent use cases (supply chain, trading) are either solved by traditional RPA/workflow tools or remain unsolved problems. Fetch's success depends on finding compelling decentralized agent applications that centralized tools genuinely can't serve. The theory looks good; practice remains thin.

Recent developments (April 2026)

Following Ocean's exit, Fetch announced plans to accelerate DAO governance, aiming for full community voting by mainnet launch. But specifics on voting thresholds, treasury management, and upgrade procedures are missing.

The ASI:Create platform entered expanded beta in Q1 2026 with better developer tools and enterprise integrations. Early testers say the IDE is easier, but real-world deployments are still rare.

ASI:Chain testnet (originally Q1 2026) got pushed to Q3 2026 due to consensus refinements and audits. The delay raised questions about timeline feasibility.

Token rebranding from FET to ASI got postponed pending governance clarity. Current estimate is Q4 2026 or later.

FAQ

A: Smart contracts are passive—they wait for a transaction to trigger them. Agents are proactive. They initiate transactions, negotiate, and adapt to changes without someone hitting a button. Agents use machine learning; contracts execute fixed logic.

A: Agents provide services (computation, data validation, market-making) in Fetch's marketplaces. They can stake as validators. Earned FET transfers to wallets or buys other agent services.

A: No. It's in testnet development (Q3 2026 target). Current Fetch mainnet is production-grade but limited to 50 TPS. ASI:Chain is experimental.

A: Ocean withdrew in October 2025 over treasury and token allocation disputes. Legal action followed. The breakdown exposed weaknesses in the alliance structure and hurt investor confidence.

A: FET depends on successful ASI:Chain launch, regulatory clarity, enterprise adoption materializing, and governance stability post-merger. Current price ($0.23) reflects skepticism. Upside exists if those risks resolve, but they're material.

A: Yes. You stake FET, run node infrastructure, maintain 24/7 availability. Validators earn block rewards and transaction fees. Delegation options exist for passive holders.

A: Fetch emphasizes real-time agent coordination. Bittensor focuses on AI quality assessment via subnets. SingularityNET targets AI service discovery. Fetch's edge is agent-native infrastructure; that advantage shrinks as EVM chains improve performance.

A: Yes. FET continues; a rebranding to ASI is planned for late 2026/early 2027 pending DAO approval. Existing FET converts 1:1 to ASI.

Related articles

• Bittensor (TAO): Decentralized AI Network Infrastructure
• Celestia (TIA): Modular Data Availability and Rollup-Centric Scaling
• Render Network (RENDER): Decentralized GPU Computing for AI and 3D Rendering
• Autonomous agents and decentralized finance: Market design and incentives
• Layer 1 blockchains: Consensus mechanisms and trade-offs
• Cosmos ecosystem: IBC and cross-chain interoperability