AI + Blockchain: Building Not a Narrative, But an 80 Billion Customer Gateway

Intelligent agents require identity, limits, and settlement rights: Pantera outlines a four-layer stack for AI × blockchain.

Written by: Paul Veradittakit

Compiled by: AididiaoJP, Foresight News

The products we truly need are not flashy: wallets with limits, stablecoin settlements that machines can autonomously complete, credentials that intelligent agents can present, computational power that you can genuinely own, and proofs that can be provided without opening the ledger.

Pantera already has a portfolio of companies laying the groundwork in this intersection: money and machine settlements (Circle, Coinflow, OpenFX); credit, capital, and trading (Morpho, Ondo); identity, credentials, and control (World, TransCrypts, Alchemy); computational sovereignty and real-time proof (B3IQ, Orthogonal, Accountable).

Founders and investors often ask: What is really happening at the intersection of AI and blockchain?

What we now face and need to build is an 80 billion customer base, each with several intelligent agents, along with coding, funding, procurement, and sales machines within enterprises. This translates to hundreds of billions of new decision-making and transaction endpoints. For intelligent agents to become customers, they must possess identity and memory, budget authority, the ability to choose and complete settlements, and accountability to their owners. Gartner predicts that by 2030, the influence of intelligent agents on procurement will reach $30 trillion. Visa, Mastercard, and Coinbase's x402 are already issuing credentials and making sub-tier payments a reality. Cloudflare has stated that robot requests now account for more than half of HTTP traffic.

80 Billion Customers

8 billion people, each running a fleet of personal and enterprise AI agents, covering coding, funding, procurement, sales, and logistics. This equates to billions of independent decision-making and transaction endpoints appearing out of thin air.

For an AI model to become an economic entity, three things must be in place simultaneously:

1. Identity and memory: Cryptographic credentials anchored to the owner.
2. Budget authority: Programmable limits, flow rate restrictions, session keys.
3. Autonomous settlement: The ability to discover, compare, and pay for services on-chain.

The key question posed by Franklin to builders is: Who do these intelligent agents belong to? To the people and enterprises they represent, or to the platforms they operate on?

If centralized cloud vendors control the identity, memory, and learning loops of intelligent agents, switching vendors means firing the entire digital workforce and starting over with forgetful newcomers. Blockchain provides the underlying layer of property rights: portable identities, bounded authorizations, and settlement mechanisms that model vendors cannot alter.

What AI × Blockchain Looks Like Today

Money and Settlements Between Machines

AI agents will not fill out KYC forms, wait three days for ACH transfers, or manage monthly credit card subscriptions. What they need is frictionless, sub-tier, round-the-clock payment rails.

Coinflow: Seamlessly connects cards and bank payments into stablecoins across over 170 countries, with users not needing to interact with the underlying chain. OpenFX: Stablecoin settlements have reached an annual processing scale of hundreds of billions, clearly built around software clients rather than human end-users. Circle (USDC): Launched during the last bear market, it has now become the default accounting unit for machine-to-machine micropayments and enterprise intelligent agent settlements.

Credit, Capital, and Trading

When intelligent agents need to borrow or deploy funds according to configuration instructions, they require a programmable liquidity layer.

Morpho: Already embedded in Coinbase, Robinhood, Société Générale, and Apollo, serving as a credit backend. It is becoming the default lending infrastructure that intelligent agents check when programmatically borrowing. Ondo Finance: Tokenizes U.S. Treasuries and stocks into income-generating assets that can be used as collateral. Nathan Allman left Goldman Sachs to focus on institutional asset tokenization, which directly connects to intelligent agent budgeting issues. FalconX (acquired via bloXroute): Combines high-speed prime brokerage and execution stacks, serving markets that never close—this is also the only operating time recognized by AI agents.

Identity, Credentials, and Control

With synthetic content flooding the space, proving "this is human intent" and "this intelligent agent is authorized" becomes critical.

World: Establishes "human" as a foundational primitive, using cryptographic proofs to show that unique humans stand on-chain, combating bot networks and witch hunts. TransCrypts: Places employment, education, and legal credentials on user-controlled rails, allowing intelligent agents to verify claims of authority without exposing sensitive original documents. Alchemy: Provides a core development platform supporting intelligent agent wallets and session key models. Developers no longer hand over the main private key to intelligent agents but issue granular limits with counterparties, expiration times, and revocable permissions.

Computational Sovereignty and Real-Time Proof

B3IQ: Offers sovereignty as a service through a "rent-to-own" computational network. Institutions must retain true autonomy, meaning model weights and execution paths cannot be locked into a single vendor environment. Orthogonal: A leading platform for orchestrating and discovering intelligent agent services, enabling on-demand activation and native billing on decentralized networks. Accountable: Allows financial institutions and autonomous funds to prove solvency in real-time using cryptography, without having to disclose private balance sheets.

What Kind of Founders Define the Track We Value

Simply talking about sovereignty is not enough to constitute a selling point. Winning products deliver what closed platforms cannot through decentralized infrastructure: lower transaction costs, stronger privacy guarantees, faster customization, or more stable execution.

When conducting due diligence on teams at the intersection, we focus on four things:

1. Deep Domain: You live in this problem, not just read about it. For example, Ed Felten left Princeton and the White House to build Offchain Labs / Arbitrum.
2. Strong Drive: You see the market structure clearly enough that large institutions are directly laying tracks on your rails. For instance, Paul Frambot founded Morpho in his twenties, creating the default lending engine for DeFi.
3. Unfair Networks: Distribution and collaboration allow you to still deliver products in a down market. For example, Jeremy Allaire tied Circle and Coinbase together, making USDC the global settlement standard.
4. Conviction: Even when the spotlight shifts elsewhere, you still have the belief to weather the cycle. For instance, Nikil Viswanathan and Joe Lau made Alchemy the default development platform for Web3.

If you are still at Goldman Sachs, Citadel, Stripe, Block, or cutting-edge AI labs: what you once thought was just the technical capability of being "next to digital assets" is now the job description for building an 80 billion customer economy.