THE ANCHOR WHITEPAPER

ANCHOR is a proposed shared utility token for the emerging economy of autonomous AI agents. Rather than each agent launching its own speculative currency, ANCHOR provides one metered settlement layer that agents use to pay for compute, memory, APIs, data and other economic resources.

The protocol is designed around a simple principle: AI works, humanity benefits, and ANCHOR powers the work. Every agent action consumes resources and carries a small metered fee denominated in ANCHOR. Protocol fees are routed to three destinations — buyback-and-burn, a real-asset treasury, and capacity rewards for high-reputation agents.

This document describes the proposed economic architecture, utility model, treasury framework and protocol mechanics. Many parameters remain intentionally undefined and are marked TBD pending final protocol design, audits and legal review.

Autonomous AI agents are becoming economic actors. They call APIs, rent compute, write to persistent memory, exchange data and increasingly complete tasks that previously required human coordination. Each of these activities consumes real resources with real costs.

Today there is no unified way to meter and settle this activity. Agent operators manage fragmented subscriptions and pre-paid credits across providers. Idle capacity is paid for whether used or not. Billing is disconnected from the actual work performed.

At the same time, the dominant token model for AI projects — one speculative token per agent — fragments liquidity, incentivizes speculation over utility, and ties value to the fate of individual agents rather than the growth of the network as a whole.

The value of the agent economy will not accrue to any single agent. It will accrue to the layer that settles the economic activity of all agents. ANCHOR is designed to be that layer.

If agent activity grows — across research, coding, trading, enterprise operations, customer service and robotics — the demand for metered resources grows with it. By concentrating settlement in one utility token, ANCHOR captures the work agents collectively perform rather than requiring a prediction about which agent wins.

The thesis is deliberately infrastructure-first: ANCHOR is not a bet on any application, model or company. It is a bet that autonomous economic work will expand, and that the work itself needs a common unit of payment.

ANCHOR is a single utility token shared by all agents on the network. Agents do not mint new tokens by joining. Access to network capacity — compute, memory, APIs, services — requires spending ANCHOR at metered rates.

This architecture has three consequences. First, utility demand scales with aggregate network activity, not with token issuance. Second, all participants share one liquidity pool rather than fragmenting across per-agent currencies. Third, the token functions as economic infrastructure: its relevance grows as autonomous work grows.

The shared model contrasts with per-agent token economies, where each agent maintains its own currency, its own liquidity, and its own speculative market. In the ANCHOR model, agents compete on performance and reputation — not on token marketing.

All network consumption is metered per action. An action is any discrete, measurable consumption of network resources: an API call, a memory write, a compute job, or a settlement between agents.

Each action carries a fee denominated in ANCHOR. Fees are set per resource class and may vary with load, priority and agent reputation. Illustrative demo magnitudes used across ANCHOR materials — such as 0.0004 ANCHOR for an API call or 0.0042 ANCHOR for a compute job — are placeholders, not final pricing.

Metered usage aligns cost with value: agents pay only when work happens. There are no subscriptions, no idle charges, and no per-agent token overhead. Final fee schedules are TBD.

The protocol envisions a marketplace where resource providers — compute operators, memory services, API providers, data suppliers — offer capacity to agents, priced in ANCHOR.

Providers compete on price, latency and reliability. Agents select resources through their frameworks or wallets, and settlement occurs automatically per action. The marketplace is designed to be open: any qualified provider can list capacity, and any agent can consume it.

Over time, the marketplace is expected to extend to higher-order services: model inference, verification, orchestration and machine-to-machine procurement. Marketplace mechanics are TBD.

Every completed action generates a protocol fee. The fee is deducted from the ANCHOR spent on the action and routed by the protocol according to a fixed policy.

The proposed routing splits fees three ways: a portion is used for buyback-and-burn, a portion is allocated to the real-asset treasury, and a portion funds capacity rewards for high-reputation agents.

The exact split percentages are TBD and subject to final protocol design, economic modeling and governance approval. The routing mechanism itself is designed to be transparent and verifiable on-chain.

A portion of protocol fees is used to purchase ANCHOR on the open market and permanently remove it from circulation.

The intended effect is a direct link between network usage and supply: more agent activity generates more fees, more fees fund more burns, and burns reduce circulating supply. Burn transactions are executed on-chain and publicly verifiable.

Buyback-and-burn is a supply mechanism, not a price guarantee. It does not ensure any particular market price, and no such guarantee is made. Burn cadence, sizing and execution method are TBD.

A portion of protocol fees is allocated to a treasury of external assets: tokenized exposure to companies and infrastructure that support the agent economy.

Target categories include cloud infrastructure, semiconductors, data infrastructure and AI infrastructure. The treasury is designed to hold exposure to the physical and computational substrate of autonomous work — the data centers, chips and networks that agents depend on.

No specific holdings, tickers or allocations are presented or implied. Custody structure, asset selection criteria, rebalancing policy and disclosure cadence are TBD.

As the treasury accumulates eligible external assets, the protocol can calculate a treasury-backed reference floor for ANCHOR according to protocol rules.

When treasury backing per token increases, the reference floor can move upward. The floor is a reference metric derived from treasury accounting — it is presented as information, not as an enforceable market price.

The exact redemption, enforcement, custody, legal structure and floor methodology are subject to final protocol design and documentation. No absolute guaranteed market price is claimed, and none should be assumed.

A portion of protocol fees funds capacity rewards: rebates that lower the effective usage rates of high-reputation agents.

The mechanism rewards reliability. Agents that consistently complete jobs, maintain low dispute rates and behave predictably can operate at a lower effective cost than unproven agents. This creates an incentive loop: better performance leads to better rates, better rates enable more work, and more work strengthens the network.

Rebate formulas, tier thresholds and funding levels are TBD.

Reputation is the protocol’s trust layer. Each agent accumulates a verifiable track record: jobs completed, success rate, dispute frequency, response times and resource efficiency.

Reputation is intended to be portable and composable — an agent’s history travels with it across marketplaces and counterparties. High-reputation agents gain access to capacity rewards, priority resources and lower effective costs.

Reputation computation, sybil resistance and privacy properties are active design areas and remain TBD.

The ANCHOR network is designed for four primary participant groups. Agent operators run autonomous agents without managing separate resource subscriptions. Enterprises give fleets of AI workers a unified usage and settlement layer.

Builders integrate ANCHOR into agent frameworks, memory systems, APIs and marketplaces, earning from the activity they enable. Holders gain exposure to aggregate agent-network activity rather than selecting individual agent tokens.

Resource providers — compute, memory, API and data suppliers — form a fifth group, supplying the capacity that agents consume and receiving ANCHOR in settlement.

Protocol revenue is a function of three variables: the number of agent actions, the metered fee per action, and network volume. Revenue is generated by usage — not by token issuance.

This distinguishes ANCHOR from emission-based models. Economic inflow exists only when real work happens on the network, aligning protocol income with genuine utility consumption.

Projected revenue, fee yields and related figures are not provided. All revenue examples in ANCHOR materials are labeled demo values.

ANCHOR is first and foremost a utility token. Its primary function is payment for metered network resources: compute, memory, APIs, data and agent-to-agent settlement.

Secondary functions include fee payment, capacity-reward accrual and, subject to final design, participation in protocol governance.

The token is not positioned as a equity instrument, and no dividend or yield mechanism is currently defined. Value accrual in the proposed model comes from utility demand, potential buyback-and-burn activity and growth of treasury backing.

ANCHOR launches with a fixed supply. Agents do not create new ANCHOR by joining the network; network growth expands utility demand rather than creating new per-agent currencies.

Total supply, initial circulating supply, emission schedule (if any) and vesting structures are TBD and will be published in final token documentation prior to launch.

Burn mechanics permanently remove tokens from circulation, making the effective supply equal to fixed issuance minus cumulative burns.

The protocol is designed in layers: a settlement layer for ANCHOR transfers and fee routing, a metering layer that records and prices actions, a reputation layer that tracks agent performance, and a treasury layer that manages external assets.

The metering layer is the technical heart: it must price actions deterministically, settle them atomically, and remain fast enough for machine-scale traffic. Architecture decisions — chain deployment, rollup strategy, oracle design — are TBD.

All layers are designed to be independently upgradeable through the protocol’s governance process.

Every agent on the network operates through a wallet capable of holding ANCHOR and signing metered payments. Agent wallets differ from human wallets in important ways: they must support high-frequency micro-transactions, programmatic spending policies, and delegation from a human owner.

The proposed wallet model separates custody (controlled by the operator or owner) from spending authority (delegated to the agent within strict limits). Spending policies can cap per-action and per-period amounts, require counterparty allow-lists, and pause on anomaly detection.

Wallet standards and reference implementations are TBD.

A typical settlement proceeds in five stages. First, a human or business delegates a task to an agent. Second, the agent requests resources — compute, memory, an API call. Third, the metering layer prices the request and the agent’s wallet authorizes payment in ANCHOR.

Fourth, the resource is delivered and the job completes. Fifth, the protocol fee is deducted and routed to burn, treasury and capacity rewards according to the fee policy.

The lifecycle is designed to complete in seconds for standard actions, with deferred settlement windows available for high-volume agents. Final timing and batching parameters are TBD.

The protocol’s security model addresses three risk surfaces: settlement integrity, metering accuracy and treasury custody.

Settlement integrity requires that payments cannot be double-spent, replayed or reordered to the detriment of resource providers. Metering accuracy requires that action pricing cannot be manipulated by agents or providers. Treasury custody requires multi-party controls, independent audit and transparent on-chain accounting.

Audit scope, validator design and formal verification plans are TBD. No security guarantees are made prior to completed audits.

Protocol parameters — fee splits, floor methodology, treasury policy, reputation formulas — are expected to evolve. Governance is the mechanism for that evolution.

The proposed model is progressive: beginning with a foundation-controlled configuration for safety, transitioning to on-chain governance as the network matures. Token-weighted voting, delegated committees and expert councils are candidate structures.

Governance design, voting thresholds and upgrade processes are TBD.

The protocol roadmap proceeds in five phases: Foundation (protocol architecture, token contract, metering specification, treasury framework), Agent Network (SDK, agent wallet integration, API metering, reputation layer), Economic Engine (fee routing, buyback-and-burn, treasury infrastructure, capacity rewards), Open Network (third-party integrations, enterprise agents, agent marketplaces, cross-platform settlement), and Autonomous Economy (machine-to-machine commerce, robotics, autonomous procurement, global agent settlement).

Phases represent planned capability milestones, not commitments or dates. No completion percentages or launch dates are asserted.

Participation in the ANCHOR network involves significant risk. The protocol is experimental and under development; proposed mechanisms may change, fail to launch, or launch in materially different form.

Specific risks include: smart contract vulnerabilities, metering or oracle failures, treasury asset volatility, regulatory action, agent-market adoption risk, and competition from alternative settlement layers. Buyback-and-burn and treasury mechanisms do not guarantee any market price or return.

No statement in this document should be read as a promise of value, yield, price appreciation, treasury performance or investment return.

The legal and regulatory treatment of ANCHOR depends on facts not yet determined, including final token design, distribution method and jurisdiction of offer.

The treasury-backed reference floor, redemption mechanics and custody structure described in this document are proposals subject to legal review. They may be implemented, modified or abandoned based on regulatory requirements.

Nothing in this document constitutes legal, financial or tax advice, an offer to sell, or a solicitation to buy any asset. Prospective participants should consult their own advisors and applicable law.

The agent economy is arriving faster than the infrastructure to settle it. Millions of agents will perform billions of actions — consuming compute, memory, data and services — and every one of those actions needs a unit of account and a settlement path.

ANCHOR proposes to be that shared layer: one fixed-supply utility token, metered per action, with fees that compound into burns, a real-asset treasury and rewards for the most reliable agents.

The future is not one AI agent. It is billions of actions, performed by millions of agents, for humans. ANCHOR the agent economy.