The Economics of AI Agent Access: Who Controls the Next Digital Transaction?

For years, digital businesses have paid to attract customers through search engines, advertising platforms and online marketplaces. Once a potential buyer arrived, the merchant could present its products, manage the checkout process and develop a direct commercial relationship. But a new intermediary is entering this model: the AI agent, capable of researching products, comparing offers and potentially initiating transactions on behalf of consumers.

The rise of agentic commerce introduces a fundamental economic question. If customers increasingly rely on AI assistants to decide where and what to buy, who controls the transaction—and who captures the value it creates? Retailers may gain access to new audiences, but they could also face new distribution costs, platform dependencies and reduced visibility into customer behavior. Meanwhile, publishers and digital service providers are beginning to explore whether automated access to their content and infrastructure should become a paid service.

By October 2026, developments from Google, OpenAI and Cloudflare have made AI agent access economics a concrete business discussion rather than a distant possibility. New commerce protocols, merchant integrations and experimental content-licensing models suggest that the next stage of digital business may depend on more than attracting website traffic. Companies must understand how AI intermediaries could redistribute revenues, operating costs, bargaining power and control over the customer relationship.

AI Agent Access Economics: From Website Traffic to Transaction Control

The traditional internet economy relies heavily on intermediaries.

Search engines organize information and direct users toward websites. Advertising platforms sell access to audiences. Marketplaces connect sellers with buyers and often charge fees for the transactions they facilitate.

These models differ, but they share a basic commercial principle: an intermediary creates value by making discovery, distribution or purchasing more efficient.

AI agents introduce another possible layer.

Instead of visiting several retailers, a customer may ask an assistant to compare options and identify the most suitable offer.

The assistant could gather information, evaluate alternatives, interact with participating merchants and potentially help complete a purchase.

The economic significance lies in the assistant’s position between the buyer and the seller.

A merchant may still provide the product, manage inventory and fulfill the order. Yet an external AI platform could influence which merchant is selected, how the offer is presented and which transaction route the customer follows.

This raises three distinct questions:

• Who controls access to the customer?
• Who receives payment for facilitating the transaction?
• Who retains the commercial relationship after the purchase?

These questions do not have universal answers.

Different companies are experimenting with different technical and commercial arrangements. However, the decisions being made now could shape the bargaining position of retailers, publishers and digital platforms.

Why the AI Agent Could Become a New Digital Gatekeeper

A gatekeeper does not necessarily manufacture products or provide the underlying service.

Its power comes from controlling an important point of access.

Online marketplaces illustrate this principle. Sellers may benefit from a platform’s established customer base, payment infrastructure and reputation.

In exchange, they accept specific participation rules and, depending on the platform, transaction fees or other commercial charges.

AI-mediated commerce could introduce a related form of dependency.

A customer may increasingly rely on an assistant to identify suitable products before visiting any merchant.

In that situation, the AI platform influences the selection process even if the final payment occurs elsewhere.

Control Over Discovery Can Create Economic Leverage

Consider two retailers offering comparable products.

The first attracts customers through its own website and existing customer relationships.

The second relies heavily on referrals from a major AI shopping assistant.

If the AI platform changes its integration requirements, recommendation mechanisms or commercial policies, the second retailer may experience greater disruption.

The risk resembles platform dependency, but the mechanism is different.

The AI intermediary may influence the customer’s choice before a conventional referral or marketplace transaction occurs.

A retailer could therefore become commercially dependent on a platform that does not directly process most of its sales.

This dependence may be difficult to measure because part of the buying journey takes place outside the retailer’s analytics environment.

The broader problem has already been examined in Market Insiders “Discovery Risk: What Businesses Lose When Customer Research Moves Outside Their Website”.

That analysis focuses on the loss of visibility into early customer research. Agentic commerce extends the question into the transaction itself: what happens when the intermediary that influences discovery also participates in how the purchase is completed?

Gatekeeping Does Not Automatically Mean Monopoly

It would be premature to conclude that one AI company will control the entire market for automated shopping.

Consumers may use different assistants, retailers may support several integrations and businesses may continue attracting customers through their own channels.

The degree of market power will depend on adoption, switching costs, commercial agreements and the ability of merchants to reach consumers through alternative services.

Nevertheless, the possibility of intermediary concentration deserves attention.

A company that controls an influential shopping interface could eventually acquire bargaining power over participating merchants, even without owning their inventories or physical distribution networks.

Google’s Universal Commerce Protocol: Who Controls the Checkout?

One of the most significant developments in 2026 came from Google.

On January 11, the company introduced the Universal Commerce Protocol (UCP), an open-source standard developed with participation from retailers and technology partners.

According to Google’s official technical explanation, the protocol establishes common mechanisms through which AI-powered consumer interfaces, merchants and payment providers can communicate.

Its purpose is to support commerce activities such as product discovery, checkout and related services through standardized integrations.

From an economic perspective, the important question is not simply whether the protocol makes transactions easier.

It is how responsibilities and commercial relationships are distributed among the participating companies.

The Merchant of Record Remains Important

Google’s merchant documentation states that participating retailers remain the seller of record within the relevant UCP-powered checkout experience.

This means the merchant continues to hold the commercial role associated with selling the product, even when the customer initiates the purchasing process through Google’s interface.

Under Google’s implementation, eligible users can complete checkout through supported Google Pay flows.

The arrangement is important because an AI interface does not necessarily become the seller simply because it facilitates a transaction.

However, remaining the seller of record does not mean the merchant controls every aspect of discovery, customer interaction or platform participation.

Those elements depend on the specific implementation.

Standardization Can Reduce Integration Costs

A potential advantage of UCP is reducing the need for retailers to build separate custom integrations for every participating AI service.

In a fragmented environment, businesses might need to support numerous incompatible transaction interfaces.

Each integration can create engineering costs, testing requirements, maintenance obligations and additional operational complexity.

Standardized capabilities could reduce some of that burden by allowing participating systems to communicate through common formats.

Nevertheless, a common protocol does not automatically eliminate integration costs.

Merchants may still need to maintain accurate catalogs, manage authentication, update backend systems and test supported purchasing flows.

They must also comply with the relevant commercial and technical requirements of each participating service.

Availability Remains Limited

As of October 2026, Google’s merchant documentation describes UCP-powered checkout as available to selected participating merchants.

The documentation identifies relevant eligibility in the United States, Canada and Australia.

This does not mean that every retailer in those countries has access or that the same checkout capabilities are available globally.

Businesses should distinguish between the publication of an open standard and the commercial availability of a particular implementation.

For investors and executives, that distinction matters when evaluating potential returns.

A protocol can be technically promising without yet delivering meaningful transaction volumes to every merchant.

OpenAI’s 2026 Commerce Strategy: Discovery Before Checkout

OpenAI’s shopping developments provide a different view of the emerging market.

On March 24, 2026, the company announced expanded product-discovery experiences in ChatGPT, supported by enhancements to the Agentic Commerce Protocol.

Its official announcement described integrations through which participating merchants can provide structured product information, prices and promotions.

The company also explained that its initial Instant Checkout experience had not delivered the flexibility it wanted to offer merchants.

As a result, it emphasized product discovery while allowing participating retailers to use their own checkout experiences.

This is an important commercial development.

Discovery Can Be Valuable Without Processing the Payment

A platform does not necessarily need to complete the transaction to become an influential distribution channel.

If an AI assistant helps consumers identify suitable products, compare alternatives and decide where to purchase, it may create commercial value before checkout.

That value could potentially support different monetization arrangements, including advertising, subscriptions, commercial partnerships or other services.

However, these should be understood as possible business models rather than evidence that a universal transaction fee is currently applied.

The actual commercial terms of participating services must be examined individually.

Why Merchant-Controlled Checkout Matters

For retailers, retaining their checkout environment can preserve important business capabilities.

They may continue managing payment options, loyalty programs, customer accounts, fulfillment arrangements and service policies through their established infrastructure.

At the same time, AI-driven discovery can still influence which products customers consider.

This produces an important distinction between transaction ownership and demand origination.

A merchant may process a sale directly while depending on an outside platform to introduce the customer.

That dependency can carry economic consequences even when no external company appears as the seller on the receipt.

Walmart Shows How Deeper Integrations Can Coexist

The same March announcement described a Walmart shopping experience integrated into ChatGPT.

The arrangement connects conversational discovery with a tailored Walmart environment supporting account linking, loyalty features and Walmart payment capabilities.

This illustrates an alternative to treating AI platforms and retailers as opponents competing for complete control over checkout.

A major retailer may choose to integrate more deeply with an external assistant while preserving selected elements of its existing customer infrastructure.

Such partnerships, however, may be easier for large companies with established technology teams and negotiation resources than for smaller independent merchants.

The resulting competitive effects will depend on how widely different integration options become available.

What Will AI Agents Cost Merchants?

One of the most important unanswered questions concerns the cost of acquiring transactions through AI intermediaries.

Traditional e-commerce businesses already evaluate customer acquisition cost, payment processing charges, fulfillment expenses and marketplace commissions.

Agentic commerce may introduce new forms of expenditure—or replace existing ones.

The economic effect will depend on whether an AI channel produces additional sales, changes existing customer behavior or simply moves purchases from one distribution channel to another.

Transaction Fees and Distribution Charges

A business using an external platform may encounter several possible charging models.

One arrangement could involve a fee for each transaction generated through the platform.

Another could charge for qualified referrals, product visibility, subscriptions or access to premium integrations.

Commercial agreements might combine several mechanisms.

Not every provider uses these structures, and no single fee schedule applies across the emerging agentic-commerce market.

The central business question is whether the additional distribution cost is justified by the sales it produces.

The Cost of Merchant Integrations

Even when a commerce protocol is openly available, implementing it can require business resources.

A retailer may need engineering work, catalog synchronization, inventory integration, testing, fraud controls and ongoing monitoring.

These costs should not be ignored simply because access to the technical standard is free.

Open standards can reduce certain integration barriers, but they do not guarantee that participating in every commercial ecosystem will be inexpensive.

Attribution and Measurement Costs

AI-mediated transactions also create a measurement challenge.

A customer may discover a retailer through one AI service, investigate the product elsewhere and complete the purchase through the merchant’s website.

If the business cannot reliably identify which interactions contributed to the sale, evaluating the profitability of the channel becomes difficult.

This may require investment in attribution tools, customer research and additional reporting processes.

Those expenses should be considered part of the total cost of operating an acquisition channel rather than treated as unrelated technology spending.

Hypothetical Financial Model: Can Agentic Commerce Improve Retail Margins?

The following example is entirely hypothetical.

The figures do not represent actual charges imposed by Google, OpenAI or any other agentic-commerce provider. They are used exclusively to demonstrate how a retailer could compare two distribution models.

Assume an online retailer evaluates two alternative routes for selling the same products.

Each route generates 1,000 orders with an average order value of $120.

The product mix, fulfillment obligations and customer-service requirements remain identical.

Scenario A: Conventional Digital Acquisition

The business acquires customers through its existing marketing channels and completes transactions through its online store.

The retailer generates $23,000 after the operating and acquisition costs included in the example.

This is not net accounting profit. Fixed overhead, taxes, financing costs and other expenses are excluded.

Scenario B: AI-Mediated Acquisition

Now assume the same business generates 1,000 comparable orders through a hypothetical AI commerce channel.

For the example, the retailer pays $14 per order for AI-mediated distribution and incurs an additional fixed integration cost of $5,000 for the period.

All other underlying transaction expenses remain unchanged.

Under these assumptions, the AI-mediated route produces $22,000, compared with $23,000 through conventional acquisition.

The difference is $1,000 in favor of the conventional route.

Where Is the Break-Even Point?

The simplified comparison can also identify the maximum AI distribution charge that would generate the same contribution as the conventional channel.

In Scenario A, acquisition spending totals $18,000.

In Scenario B, the retailer incurs $5,000 in additional integration expenses.

To match the conventional channel’s total acquisition-related cost, the AI distribution charges would therefore need to remain at or below:

$18,000 − $5,000 = $13,000

Dividing this amount by 1,000 orders gives a break-even distribution charge of:

$13 per order

At that hypothetical fee, both channels would generate the same $23,000 contribution after the specified costs.

Above $13, the AI-mediated route becomes less attractive under these assumptions. Below $13, it becomes more attractive.

Why the Result Changes With Scale

The $5,000 integration cost has been treated as a fixed cost for the period.

If more orders were generated through the same integration, the average fixed cost per order would fall.

Conversely, if transaction volumes were significantly lower, the integration expense would represent a larger burden per sale.

The calculation also assumes both channels generate comparable orders.

In reality, conversion quality, return rates, average order values and customer retention may differ.

A company should therefore avoid evaluating an AI commerce partnership on headline transaction fees alone.

The more relevant question is:

How much incremental contribution does the channel generate after all additional costs, compared with the realistic alternatives?

If AI-mediated orders largely replace sales the retailer would have acquired through cheaper direct channels, the new arrangement could reduce profitability even when total sales remain stable.

If the channel generates genuinely new customers at an attractive cost, it may create additional economic value.

Distinguishing incremental demand from displaced demand is essential to an accurate return-on-investment calculation.

Content Licensing: A New Revenue Model for the AI Economy

Agentic commerce does not only affect retailers.

Publishers, research organizations and information providers also face questions about how their content creates commercial value when accessed or used by AI services.

The economics are different from selling physical goods.

A publisher may generate revenue from advertising, subscriptions, licensing or direct customer relationships.

When an AI service retrieves information and presents its own response, the original publisher may not receive the website visit that would normally support those revenue models.

This creates a potential mismatch between the value of the information and the compensation received by the organization that produced it.

Cloudflare’s Pay Per Crawl

Cloudflare has developed tools that allow website owners to control access by recognized AI crawlers.

Its Pay Per Crawl system introduces an experimental model in which participating publishers can charge for successful content retrieval requests.

According to Cloudflare’s documentation, the system remained in closed beta as of July 2026.

The underlying economic idea is that retrieving content can become a billable event rather than an activity that always takes place without direct payment.

However, AI crawlers and user-directed shopping agents are not the same category of software.

Pay Per Crawl is a content-access mechanism. It should not be described as a universal transaction fee imposed on AI shopping agents.

Pay Per Use: Charging for the Value of Content

On September 30, 2026, Cloudflare announced a second experimental approach called Pay Per Use.

In its official announcement, the company described a model in which an AI service proposes a price for a defined use of publisher content.

Participating publishers can choose whether to accept the proposed terms.

The AI service then reports qualifying uses, while Cloudflare provides billing and payment infrastructure.

This is different from charging every time a crawler accesses a webpage.

The concept attempts to link compensation to downstream uses, such as content appearing in an AI-generated response or influencing an automated recommendation.

For publishers, it suggests an alternative revenue stream that may not require the consumer to visit the original website.

However, the system was still in beta at the time of the announcement.

The commercial viability of the model cannot be assumed, and the amounts publishers may ultimately earn will depend on participation, pricing and actual usage.

A New Business Risk: Who Verifies the Usage?

Cloudflare’s initial Pay Per Use system relies on buyers reporting qualifying uses.

That creates an important commercial question.

If payment depends on the AI company accurately reporting when and how it used external content, publishers need sufficient confidence in the reporting process.

Incomplete reporting could reduce compensation.

Disagreements over what qualifies as a billable use could also create disputes between content owners and AI service operators.

The reporting mechanism may evolve, but the underlying issue is familiar from other digital markets.

Reliable measurement is essential when compensation depends on activity that one commercial party can observe more directly than another.

When APIs Become Billable Services for AI Agents

Content licensing is only one possible form of monetization.

Businesses also provide structured data, software capabilities and APIs that automated systems may want to access.

For example, an AI agent might request live travel availability, financial market data, logistics information or an enterprise software function.

These services can involve operational costs that differ substantially from serving an ordinary webpage.

They may require computing resources, database queries, authentication, support and ongoing infrastructure maintenance.

Some providers may therefore choose to monetize authorized agent access through usage-based pricing.

Cloudflare’s September 2026 announcement also introduced a beta Monetization Gateway concept for charging automated clients per request.

This illustrates the possibility of creating payment infrastructure directly around machine-to-machine services.

Usage-Based Pricing Creates Opportunities and Risks

Charging per request could help service providers recover infrastructure costs and monetize valuable capabilities.

However, it introduces expenditure risks for the companies operating agents.

An agent performing one customer task might make multiple requests across different services.

If each request carries a fee, the total cost of completing the task depends on how the agent operates.

Retries, duplicated queries and inefficient workflows could increase expenditure without necessarily improving the result.

This makes cost monitoring, spending limits and service-level agreements important components of agentic operations.

For businesses, the economic challenge becomes managing the cost of automated work rather than simply measuring the number of human users.

Merchant Margins and the Risk of Fee Stacking

An important concern in digital platform economics is the accumulation of charges across the transaction process.

Consider a retailer that already pays for advertising, marketplace access, payment processing and logistics.

If an AI intermediary introduces an additional commercial fee, the retailer needs to understand whether that fee replaces an existing cost or adds another layer.

The distinction is critical.

If AI distribution reduces the need for conventional advertising, the additional charge may be commercially acceptable.

But if the business continues paying the same advertising expenses while also paying the AI intermediary, the combined cost of acquisition can increase.

Every Additional Intermediary Needs an Economic Role

A sustainable transaction model should provide identifiable value at each stage.

The AI service might reduce search friction or introduce new buyers.

A payment provider may manage payment execution and associated services.

A marketplace may provide fulfillment, customer protection or access to an established audience.

The retailer contributes inventory, product quality and delivery capabilities.

The challenge is ensuring that the total cost of these services does not exceed the economic value they create.

For low-margin products, even relatively small additional charges can materially affect contribution margins.

For higher-value or specialized products, the retailer may have greater flexibility, particularly when an intermediary generates customers who would otherwise be difficult to reach.

The relevant metric is not simply the fee charged by one participant.

It is the combined cost of acquiring, processing and fulfilling the transaction.

Platform Dependency: The Hidden Cost of AI-Mediated Distribution

The economic risks of agentic commerce extend beyond direct fees.

When a retailer depends heavily on an external AI platform, changes to that platform’s commercial rules or technical integrations could affect business performance.

The risk becomes more significant when a large share of customer acquisition is concentrated within one ecosystem.

For example, an online merchant might invest in a custom integration with a major AI assistant.

If the platform changes its technical requirements, the merchant may need additional engineering work.

If the commercial terms change, the same volume of sales may become less profitable.

If customer demand shifts toward another assistant, the original integration may no longer justify its cost.

These are potential business scenarios, not evidence that a specific platform is preparing to impose such changes.

Switching Costs Can Create Commercial Lock-In

Once a retailer has built internal workflows around a particular AI platform, moving to an alternative may become expensive.

The business may need to replace integrations, retrain staff, update data feeds and rebuild reporting systems.

These costs can create a form of commercial lock-in even without exclusive contracts.

Open interoperability standards may reduce some switching barriers, but they cannot eliminate every operational dependency.

The company’s strategic flexibility depends on how easily it can move between partners while maintaining customer access and transaction continuity.

A Multi-Platform Strategy Has Its Own Costs

Diversifying across multiple AI services may reduce dependence on a single provider.

However, every additional integration creates maintenance and coordination requirements.

A business could protect itself from concentration risk while simultaneously increasing operational complexity.

The correct strategy is therefore not necessarily to integrate with every available assistant.

It is to identify which partnerships offer measurable value and maintain credible alternatives where the economic exposure justifies them.

Who Owns the Customer Relationship After an AI-Mediated Purchase?

In digital commerce, ownership of the customer relationship is more complicated than identifying which company processes the payment.

A retailer may remain the legal seller while an AI platform manages much of the discovery experience.

The merchant may receive the order details necessary for fulfillment but have limited visibility into the conversations that preceded the sale.

The exact information available will depend on the technical integration, privacy requirements and applicable commercial arrangements.

Customer Data Is an Economic Asset

Customer information can help businesses understand repeat purchasing, product preferences, service problems and demand trends.

However, businesses must distinguish between data they are legally entitled to process and data they would simply prefer to obtain.

An AI platform may have visibility into consumer requests that are never shared with participating merchants.

A retailer may understand what was purchased without knowing every alternative the customer considered.

This creates an information asymmetry between different participants in the transaction.

The economic importance of that asymmetry depends on whether the missing information affects decisions about pricing, inventory, retention and product development.

Repeat Customers Can Change the Economics

The first transaction does not always represent the full commercial value of a customer.

A retailer may accept relatively high acquisition costs if it expects repeat purchases and has sufficient evidence that the relationship will remain profitable.

But an AI-mediated purchase may create uncertainty about which channel the customer will choose next time.

Will the consumer return directly to the merchant?

Will they ask the same AI assistant to find the best available offer again?

Or will a different intermediary influence their next purchase?

These possibilities affect customer lifetime value and the economic attractiveness of external acquisition channels.

Companies should avoid assuming that customers obtained through AI integrations will necessarily have the same retention characteristics as customers acquired through established channels.

Smaller Businesses Face a Different Competitive Challenge

Agentic commerce creates opportunities for smaller retailers to reach consumers through new discovery interfaces.

However, participation may involve operational requirements that are easier for larger organizations to absorb.

A major retailer may have established product-feed infrastructure, dedicated integration teams and resources for negotiating commercial agreements.

A smaller merchant may rely on third-party e-commerce platforms, external developers or standardized integrations.

The cost of maintaining accurate inventory, product data and technical compatibility can therefore become an important consideration.

Open Standards Could Lower Barriers to Entry

Common protocols have the potential to reduce fragmentation.

If participating AI platforms support interoperable capabilities, smaller businesses may be able to reach multiple services through fewer integrations.

That could reduce some technical disadvantages associated with scale.

Nevertheless, open technical specifications do not guarantee equal commercial visibility or identical access to every feature.

Actual participation may still depend on eligibility, platform policies, geographical availability and the capabilities of the merchant’s existing systems.

The Importance of Contribution Margin

For smaller companies, the main question should not be whether agentic commerce is fashionable or technologically impressive.

It should be whether participating in a particular channel improves economic performance.

A merchant needs to know how much revenue is genuinely incremental, what acquisition and integration costs apply, and whether the resulting orders generate sufficient contribution.

Without those measurements, a new distribution channel can increase transaction volume while weakening profitability.

What Businesses Should Measure Before Investing in Agentic Commerce

Executives evaluating AI commerce partnerships should move beyond headline announcements and examine a set of financial and operational indicators.

The most important measures include:

• Incremental revenue: Sales that would be unlikely to occur without the AI channel, rather than orders redirected from existing channels.
• Contribution margin per order: Revenue remaining after product costs, fulfillment, acquisition, transaction fees and other relevant variable expenses.
• Total cost of integration: Initial development expenses, ongoing maintenance, system updates and operational support.
• Customer acquisition cost: The combined cost of obtaining customers through the AI channel, including any overlapping marketing expenditure.
• Channel concentration: The proportion of sales or new-customer acquisition dependent on an individual intermediary.
• Repeat purchase behavior: Whether customers acquired through an AI service return directly or continue purchasing through intermediaries.
• Data and reporting access: Which commercial signals the merchant can legitimately observe and use.
• Operational risk: Failed transactions, integration outages, disputes, refunds and the business impact of service interruptions.

The importance of each metric will vary by business model.

A publisher monetizing research through AI content licensing faces different economics from a retailer selling physical products.

A software company charging agents for API calls has different cost drivers from a marketplace facilitating consumer transactions.

The common requirement is to identify precisely where value is created and how much each participant pays to obtain it.

The Real Economic Question Is Bargaining Power

The development of agentic commerce could change the balance of power between merchants, AI platforms, publishers and technology infrastructure providers.

Businesses that own distinctive products, valuable data or trusted customer relationships may retain considerable negotiating leverage.

Others may become increasingly dependent on external systems for visibility, transaction access or distribution.

The outcome will depend partly on how concentrated the market becomes and whether competing platforms provide viable alternatives.

Open standards could reduce some barriers, but technical compatibility does not automatically guarantee balanced commercial relationships.

Likewise, paid access models may create revenue opportunities for content owners, but their effectiveness will depend on participation and credible measurement.

A company that understands these dependencies can negotiate and invest more intelligently.

A company that treats agentic commerce as simply another source of traffic may underestimate the long-term implications.

How This Economic Shift Connects to the Wider Agentic Web

The economic issues surrounding AI agents are closely connected to the technical limitations of the emerging web.

As Athens Pulse explains in “The Agentic Web Hits a Wall: Why Websites Are Blocking AI Agents”, websites increasingly face questions about automated access, identity verification and the boundaries between legitimate agents and unwanted bots.

From a business perspective, access restrictions can also influence who is allowed to participate in new commercial channels and under what conditions.

The customer-facing consequences are equally significant.

In its published analysis “When AI Agents Become the Customer: The New Challenge for Brands”, Targeted.gr examines how AI intermediaries can influence product discovery, consumer trust and the relationship between businesses and buyers.

Market Insiders approaches the same technological shift from a different direction: what those interactions mean for margins, distribution costs, platform dependency and control over commercial value.

Finally, consumers must understand what they authorize AI assistants to do.

The forthcoming Techrow.gr guide, “How to Use AI Browsing Agents Safely: Permissions, Privacy and Payments“, will address the practical side of delegated online activity, including account permissions, sensitive information and transaction approval.

Together, these perspectives reveal a broader transformation: AI agents are not simply changing how the internet is used. They are creating new questions about who controls access, who accepts risk and who receives compensation when digital services interact.

Conclusion: The Next Digital Transaction May Have More Than One Gatekeeper

The emergence of AI agents creates new opportunities for retailers, publishers and digital service providers.

Automated systems may help consumers discover products, reduce some transaction friction and create additional distribution channels.

At the same time, they introduce questions about platform fees, content licensing, customer data and commercial dependencies.

Developments in 2026 demonstrate that the market is experimenting with different answers.

Google’s Universal Commerce Protocol emphasizes standardized commerce integrations while preserving the merchant’s role as seller of record.

OpenAI’s shopping strategy highlights the value of product discovery and merchant-controlled checkout.

Cloudflare’s experimental monetization tools explore ways for content owners and service providers to receive compensation for AI access and usage.

These approaches are not identical, nor have they produced one universally established commercial model.

For business leaders, the strategic priority is to evaluate agentic commerce through measurable economic outcomes rather than technological expectations alone.

An AI intermediary creates value when it brings profitable demand, reduces meaningful costs or supports transactions that would otherwise be difficult to complete.

It creates risk when new fees, integration expenses or platform dependencies grow faster than the value generated.

The defining economic question of the agentic web is not merely whether AI can complete a transaction. It is who controls the route to that transaction—and how much of its value each participant ultimately keeps.

Frequently Asked Questions About AI Agent Access Economics

What is AI agent access economics?

AI agent access economics refers to the commercial relationships, costs and revenue opportunities created when AI agents interact with websites, digital services and commerce platforms. It includes distribution fees, content licensing, integration expenses, payments and the allocation of value between businesses and intermediaries.

What is agentic commerce?

Agentic commerce describes shopping and transaction workflows in which AI systems assist users with tasks such as researching products, comparing offers and initiating purchases. The capabilities vary by provider, and many transactions still require explicit user approval or completion through participating merchants.

Who makes money from AI-mediated transactions?

Potential participants include merchants, AI platform operators, payment providers, marketplaces and other service providers. Their revenues depend on the commercial arrangements involved. There is no universal fee structure or single business model governing all AI-mediated transactions.

Do AI shopping agents charge merchants a commission?

Some commercial arrangements may include transaction fees or other charges, but there is no universal commission applied across all AI shopping systems. Businesses must examine the specific terms of each participating service rather than assuming that every agentic transaction carries the same fee.

Does Google’s Universal Commerce Protocol make Google the seller?

Not under the merchant arrangements described in Google’s UCP checkout documentation. Participating merchants remain the seller of record. However, the distribution of customer-facing interactions and platform responsibilities depends on the implementation.

Why did OpenAI change its Instant Checkout strategy in 2026?

In March 2026, OpenAI stated that its initial Instant Checkout implementation did not provide the desired flexibility for merchants. The company emphasized product discovery and allowed merchants to use their own checkout experiences while continuing to develop deeper integrations.

What is the difference between Cloudflare Pay Per Crawl and Pay Per Use?

Pay Per Crawl is designed to charge participating AI crawlers for content access. Pay Per Use is an experimental model that associates payment with specified downstream uses of publisher content. Both represent developing approaches to monetization and should not be confused with universal shopping-agent transaction fees.

Can publishers earn money when AI assistants use their content?

Potentially. Licensing agreements and experimental programs such as Cloudflare Pay Per Use provide mechanisms for compensation. However, participation, usage reporting and pricing determine whether a publisher actually receives revenue, and no universal payment arrangement applies to all AI-generated answers.

How does agentic commerce affect customer acquisition cost?

It may introduce a new distribution cost or replace part of existing marketing spending. The actual effect depends on the fees, integration expenses and additional sales generated. Businesses should compare total acquisition costs and contribution margins rather than evaluating a single platform charge in isolation.

What is the risk of depending on one AI commerce platform?

A business may become exposed to changes in platform rules, integrations, commercial terms or customer demand. The risk increases when a large share of acquisition or sales depends on one intermediary and the company lacks practical alternatives.

Is agentic commerce profitable for small businesses?

It can be, but profitability is not guaranteed. Smaller merchants need to consider implementation costs, platform eligibility, contribution margins and whether AI-mediated transactions represent additional demand or simply replace existing sales. Technical availability alone does not establish a viable business case.

What should companies measure before investing in AI commerce?

Businesses should evaluate incremental revenue, contribution margin, customer acquisition cost, integration and maintenance expenses, repeat purchasing, channel concentration and the quality of available reporting. These indicators provide a more meaningful assessment than transaction volume alone.

Will AI agents replace online marketplaces?

There is no evidence that all marketplaces will be replaced. Agentic commerce may create additional channels, deepen existing partnerships or encourage marketplaces to develop their own AI capabilities. Several business models are likely to coexist as the technology and market evolve.