AI Agents Processed 138 Million Crypto Payments in a Year, Most Worth Just a Little

Author Name: John
Author Bio: John is a cryptocurrency and financial news writer covering digital assets, blockchain technology, artificial intelligence, crypto payments, and emerging trends in the global digital economy.
Publish Date: October 15, 2026
Updated Date: October 15, 2026

Artificial intelligence is beginning to play a growing role in digital payments, with AI agents processing more than 138 million cryptocurrency payments through the x402 protocol on Coinbase’s Base network. According to a new report from blockchain analytics firm Chainalysis, most of these transactions involved extremely small amounts, demonstrating how cryptocurrency can support automated payments for online services.

The report highlights the growing connection between artificial intelligence, stablecoins, and blockchain infrastructure. Rather than relying on people to approve every purchase, AI agents can use pre-funded wallets to pay for computing resources, online data, software tools, and other digital services while completing tasks.

However, the rising transaction count does not necessarily mean that every payment represents genuine independent activity by an autonomous AI system. Researchers continue to examine how much of this activity reflects practical commercial demand.

AI Agents Record More Than 138 Million Crypto Payments

Chainalysis reported that cumulative payments through the x402 protocol on Base increased from approximately 1,500 in mid-2025 to more than 138 million by the third quarter of 2026.

The growth reflects the emergence of agentic commerce, a model in which AI systems can purchase digital services and exchange value through automated processes.

Developers can provide AI agents with wallets funded with cryptocurrency or stablecoins. When an agent needs information or a particular service, it can initiate a payment under predefined permissions and receive the requested resource.

For example, an AI assistant working on a research task might pay for a specialised data query, access a premium API, or purchase additional computing capacity without requiring a person to approve each individual transaction.

The x402 protocol helps facilitate these interactions by providing a payment mechanism for internet-based services.

Why Are Most AI Crypto Payments Worth Only a Few Cents?

One of the most notable findings is the small size of individual payments.

According to Chainalysis, more than 87% of tracked x402 transactions during the report’s observation period were worth between one-hundredth of a cent and 10 cents.

These payments are known as micropayments because their value is extremely small compared with conventional consumer or business transactions.

Such low-value payments can be difficult to process economically through traditional card networks because fixed transaction fees may exceed the cost of the service itself.

Blockchain-based payments using stablecoins can make these transactions more practical, particularly when network fees are low.

For AI agents that repeatedly access online resources, even a payment of a few cents can cover a single search, data request, software operation, or model-inference task.

The result is a payment model designed around frequent, automated transactions rather than occasional purchases made directly by human users.

What Is the x402 Payment Protocol?

The x402 protocol is an internet-native payment system introduced by Coinbase in 2025. It uses the HTTP 402 payment-required status code to help online services request payment before providing access to a resource.

The protocol can support applications in which software agents pay for services automatically, subject to the permissions and controls established by their developers.

Its potential applications include:

• Paying for API requests and real-time data.
• Purchasing computing resources.
• Accessing AI models and specialised software tools.
• Paying for code analysis and security checks.
• Enabling automated transactions between software services.

Base, Coinbase’s Ethereum layer-2 network, has become a major environment for x402 payment activity.

However, the reported 138 million transactions relate to activity tracked through this particular protocol and network. They should not be interpreted as the total number of all AI-related cryptocurrency payments worldwide.

Stablecoins Support Automated Machine-to-Machine Payments

Stablecoins are particularly useful for automated payments because their values are generally designed to track a reference asset, such as the US dollar.

USD Coin (USDC), issued by Circle, is one of the stablecoins used in the x402 ecosystem.

For developers, stablecoins offer a way to give AI agents predictable spending balances without requiring them to manage the price volatility associated with assets such as Bitcoin.

This can be useful when an agent must make repeated small purchases while staying within a predetermined budget.

For example, a business could fund an AI agent with a limited amount of USDC and allow it to pay for approved data services as needed. Spending limits and wallet permissions can help restrict how the funds are used.

Nevertheless, stablecoins are not risk-free. Users and businesses must consider issuer and counterparty risks, blockchain security, wallet management, and the possibility of incorrect or unauthorised transactions.

The United States Leads Tracked AI Model Payments

The Chainalysis report also identified a strong concentration of AI-related payment activity associated with the United States.

US-linked wallets accounted for approximately 88.1% of geographically identifiable payments for AI model inference recorded through the protocol.

AI model inference refers to the process of using a trained AI model to generate an answer, prediction, or other output. Developers may pay for these services when an agent needs additional reasoning, specialised capabilities, or access to a different model.

Chinese-linked wallets represented a much smaller share of identifiable model-inference payments, although they accounted for a substantial proportion of payments for security and code-analysis services in the report’s tracked categories.

These findings describe activity visible through the x402 system and wallets that could be geographically attributed. They do not represent the entire global AI market or all transactions conducted by AI agents.

Are AI Agents Becoming Independent Economic Participants?

The growing number of automated payments raises an important question: can AI agents become meaningful participants in the digital economy?

The current infrastructure allows developers to give software systems wallets, limited budgets, and permission to purchase services when required.

This creates opportunities for automated digital commerce, including systems that buy data, pay for computing resources, and interact with other software without requiring a person to authorise every transaction.

Chainalysis also identified signs that some wallets are being used more consistently. The median time between a wallet receiving its first deposit and making its first payment reportedly fell from 73 days in late 2025 to 18 minutes by the third quarter of 2026.

The share of qualifying wallets receiving another deposit within 30 days also increased, suggesting that some developers are funding ongoing activity rather than conducting only one-off experiments.

However, transaction data alone cannot establish how independently an agent operates or whether every recorded payment corresponds to a genuine commercial service. Some activity may come from scripts, testing, or transactions that do not reflect sustained end-user demand.

What Does This Mean for the Cryptocurrency Industry?

The growth of AI-driven payments could create new demand for blockchain infrastructure, especially networks that support fast and inexpensive transactions.

Potential implications include:

More micropayment activity: AI agents could generate large numbers of low-value transactions as they access paid services.

Greater stablecoin usage: Dollar-linked tokens may become useful for budgeting and settling automated online purchases.

New demand for payment infrastructure: Developers may need wallet services, transaction monitoring, spending controls, and tools for managing automated payments.

New security challenges: AI agents require safeguards against malicious instructions, compromised credentials, unauthorised spending, and poorly configured permissions.

More competition among payment providers: Crypto companies, card networks, and technology firms are developing systems designed to support payments initiated by AI agents.

The longer-term impact will depend on whether businesses find these systems cheaper, faster, and more reliable than existing payment options.

Challenges Facing AI-Powered Crypto Payments

Despite the rapid growth in recorded transactions, several challenges remain.

First, developers must ensure that AI agents cannot spend beyond their intended budgets or be manipulated into making unauthorised payments.

Second, businesses need reliable methods to verify that a payment has been completed before providing a service.

Third, regulators and financial institutions may need to address questions involving automated transactions, consumer protection, fraud prevention, and compliance.

Finally, the industry must distinguish genuine economic activity from automated testing or repetitive transactions that inflate headline figures without creating comparable commercial value.

These issues will influence whether agentic commerce develops into a widely used payment model or remains concentrated among early adopters and experimental applications.

Final Thoughts

The processing of more than 138 million cryptocurrency payments through x402 highlights the potential of blockchain technology to support automated digital commerce. Most of the tracked transactions were worth only a few cents, making low-cost settlement particularly important.

Stablecoins, programmable wallets, and internet-native payment protocols are creating ways for AI agents to purchase data, computing power, and software services as they work.

Nevertheless, high transaction volumes alone do not prove that autonomous AI commerce has reached mainstream adoption. Security, transaction quality, genuine customer demand, and economic sustainability will remain important measures of progress.

The key development is that cryptocurrency is becoming a practical payment option for certain machine-to-machine transactions that are too small or frequent for conventional payment methods. Whether this trend expands into a major source of crypto demand will depend on real-world use, dependable infrastructure, and stronger safeguards for automated spending.