Machine-to-machine (M2M) payments are rapidly evolving from an experimental concept into one of the fintech industry’s newest infrastructure battlegrounds, as autonomous AI agents increasingly transact without direct human involvement.
New industry data research by Keyrock shows that AI agents have already processed more than $73 million across 176 million autonomous transactions over the past year, while major financial institutions and infrastructure providers have committed more than $8 billion toward acquisitions tied to the emerging “agentic” payments ecosystem.
The shift is accelerating the development of what many in the industry now describe as the “agentic payment stack” – a new financial infrastructure layer designed specifically for autonomous software agents, AI systems, and machine-native commerce.
Fintech firms race to build the infrastructure layer for autonomous commerce
Early discussions around machine payments largely centered on whether traditional card rails, blockchain networks, or proprietary protocols would dominate the market.
Instead, the sector is increasingly consolidating around a layered infrastructure model where multiple payment standards coexist across authentication, authorization, orchestration, settlement, and compliance.
Several major fintech and infrastructure providers are now positioning themselves as foundational players within the emerging stack.
Coinbase is advancing x402, a crypto-native payments protocol using stablecoins and programmable wallets to transform wallets into machine-readable API credentials.
Meanwhile, Stripe and Tempo are supporting MPP, an orchestration framework designed to unify traditional cards, crypto payments, and Lightning Network transactions through standardized HTTP-based requests.
Google has introduced AP2, an authorization framework allowing users to cryptographically delegate spending authority to AI agents under predefined permissions and conditions.
At the same time, Visa is extending traditional card infrastructure into AI-native transaction environments through tokenized credentials designed for autonomous commerce.
Together, these initiatives are forming the foundation of a machine-centric payment architecture rather than a single dominant protocol.
Stablecoins and Layer 2 networks address the “$0.30 wall”
One of the largest catalysts behind blockchain adoption in autonomous commerce is transaction economics.
Traditional payment rails remain poorly suited for AI-driven micropayments because fixed processing costs, often around $0.30 per transaction – make low-value automated payments economically impractical.
Current industry estimates suggest that roughly 76% of agent-driven payments are valued below $0.30, effectively disqualifying conventional payment infrastructure for large portions of machine-native commerce.
As a result, stablecoin settlement on Layer 2 blockchain networks is increasingly emerging as the preferred alternative due to dramatically lower transaction costs, in some cases as low as $0.0001 per payment. USD Coin currently dominates the segment, accounting for approximately 98.6% of machine-payment settlement volume and increasingly functioning as a de facto settlement currency for autonomous financial activity.
Infrastructure consolidation accelerates as incumbents pursue vertical integration
The rapid growth of agentic payments is also triggering consolidation across the financial infrastructure market. Major incumbents are increasingly pursuing vertical integration strategies aimed at controlling multiple layers of the emerging payment stack, including wallets, identity systems, transaction routing, settlement infrastructure, and compliance tooling.
Both Coinbase and Stripe are now estimated to operate across five of the six major infrastructure layers currently defining the market.
The consolidation trend has intensified through a wave of acquisitions over the past year, including:
- Capital One acquiring Brex for $5.15 billion,
- Mastercard purchasing BVNK for $1.8 billion,
- and Stripe acquiring Bridge for $1.1 billion.
The deals reflect growing competition to become the foundational infrastructure provider for AI-native commerce.
Governance and regulatory oversight emerge as the next major challenge
While the technical infrastructure supporting autonomous payments is maturing rapidly, governance and trust frameworks remain significantly underdeveloped.
AI agents now account for as much as 75% of peak-day activity on some blockchain ecosystems, including Gnosis, while more than 104,000 autonomous agents are reportedly registered across various marketplaces and directories. However, analysts increasingly warn that regulation may become the defining challenge for the next phase of industry growth.
The implementation of frameworks such as the European Union’s MiCA regulations and the EU AI Act is expected to increase scrutiny around fraud liability, transaction authorization, accountability, and compliance controls for non-human financial actors.
Industry observers believe adoption will likely develop from the bottom up, beginning with low-value, high-frequency micropayments before gradually expanding into larger autonomous financial activity as institutional trust and regulatory frameworks mature.
The broader race is no longer simply about enabling machine payments it is increasingly about building the compliance, security, and governance infrastructure capable of supporting fully autonomous commerce at scale.
