AI Agents Are Becoming the Next Ad Network

AI Agents Are Becoming the Next Ad Network

The next major advertising network may not look like a network at all.

It may look like a conversation.

A consumer asks an AI assistant for the best running shoes for wide feet, a sofa that fits a small apartment, or a skincare routine for sensitive skin. The AI interprets the request, compares products, explains trade-offs, surfaces an offer, and may eventually help complete the purchase.

That interaction creates a new commercial environment: one where discovery, consideration, recommendation, advertising, and transaction can happen inside the same decision flow.

AI agents are not simply another place to display ads. They are becoming a new layer between consumer intent and commerce supply.

That could reshape how advertising networks are built.

From Media Inventory to Decision Inventory

Traditional ad networks aggregate media inventory.

Publishers, apps, search engines, social platforms, and video services create pages or feeds where ads can be placed. Advertisers bid for impressions based on audiences, keywords, content, or predicted behavior.

The unit of value is usually an opportunity to show an ad.

AI environments create a different kind of inventory: an opportunity to assist a decision.

The user is not merely consuming content. They are expressing a need, constraint, preference, and level of purchase intent in natural language.

For example:

> I need a waterproof jacket for a four-day trip. It should pack small, work in cold rain, and cost under $250.

This contains more commercial context than a keyword or audience segment. The AI can understand the product category, use case, budget, weather condition, trip duration, and desired attributes at the same time.

The advertising opportunity is no longer just “show this person an outdoor apparel ad.”

It is “identify the product or offer that is genuinely useful within this decision.”

That is the shift from media inventory to decision inventory.

The Market Is Already Moving

In 2026, major AI and search platforms are testing new commercial formats built around conversational intent.

Google is introducing AI-powered formats such as Conversational Discovery ads, Highlighted Answers, and Direct Offers inside AI-assisted search experiences. Direct Offers can surface a relevant promotion when a shopper is close to purchasing, while Universal Commerce Protocol integrations can connect discovery with native checkout.

OpenAI is testing ads in ChatGPT for eligible users in the United States. Its current approach separates sponsored placements from the model's answers and emphasizes clear labeling, answer independence, privacy, and user control.

The implementations differ, but the direction is consistent:

• Advertising is entering conversational environments.
• Product feeds are becoming more important than static creative alone.
• Offers can be selected based on current decision context.
• Discovery and transaction are moving closer together.
• Trust and clear sponsorship disclosure are becoming core product requirements.

The next ad network will need to understand both language and commerce.

What Makes an AI Ad Network Different?

An AI-native advertising network changes five parts of the traditional model.

1. The intent signal becomes richer

Keywords compress intent into a few words. Cookies and audience segments infer intent from past behavior. Conversations reveal intent directly.

An AI system can interpret requirements, exclusions, urgency, budget, style, and uncertainty. It can also ask follow-up questions before presenting an option.

This creates a more precise matching problem: not just who the user is, but what they are trying to accomplish now.

2. The product becomes the unit of competition

In a conventional campaign, advertisers group products, choose audiences, and optimize aggregate performance.

In an agentic environment, the AI evaluates individual products against the user's request. Each SKU competes based on relevance, price, availability, merchant quality, shipping, product attributes, and expected transaction outcome.

The winning advertiser may not be the brand with the broadest campaign. It may be the merchant with the best product-level data and offer for that specific decision.

3. The ad unit becomes an action

Traditional ads ask the user to click and continue elsewhere.

AI-native commercial formats can do more:

• Explain why a product fits the request
• Compare it with alternatives
• Apply a relevant promotion
• Build a bundle
• Confirm availability
• Answer product questions
• Route to the right merchant
• Start or complete checkout

The ad evolves from a message into a useful commercial action.

4. Creative becomes dynamic context

Static creative still matters, but it is no longer sufficient.

An AI agent may need structured product facts, brand rules, FAQs, comparison evidence, compatible products, substitutes, promotion conditions, and fulfillment information. The commercial experience can then be assembled around the user's specific need.

In this model, product data and brand knowledge become part of the creative system.

5. Measurement moves closer to outcomes

When discovery, recommendation, offer, and transaction are connected, the system can learn from more than clicks.

It can observe:

• Whether the recommendation was accepted or rejected
• Which comparison changed the decision
• Whether an offer closed the sale
• Whether the order was cancelled or returned
• Whether the customer purchased again
• Which merchant, publisher, creator, or product created incremental value

This creates the possibility of optimizing for successful commerce rather than media engagement alone.

The New Supply Side: Merchants, Publishers, and Creators

AI assistants need more than advertiser budgets. They need trusted commerce supply.

Merchants provide product truth

Merchants supply catalogs, inventory, prices, promotions, shipping, returns, and transaction outcomes. Without accurate and current product data, an AI agent cannot make reliable recommendations.

Google's expansion of Merchant Center attributes for conversational commerce reflects this requirement. Product data must increasingly describe not only what an item is, but also questions, compatibility, alternatives, and real-world use cases.

Publishers provide authority and context

Publishers contribute reviews, expertise, comparison content, and category knowledge. AI systems need credible sources to explain why a product fits a decision, not only a feed that says the product exists.

In an AI ad network, publisher content can become part of the reasoning and validation layer, while commerce infrastructure connects that influence to measurable outcomes.

Creators provide trust and culture

Creators show how products look, feel, and perform in real life. They translate product information into social proof and community relevance.

AI can help match creator content with the right shopper intent, but it should preserve attribution and creator economics. Otherwise, the network captures the value of influence without rewarding its source.

The strongest AI commerce network will connect merchant supply, publisher authority, creator influence, and consumer intent rather than replacing one participant with another.

Trust Is the Constraint, Not a Feature

AI assistants are used for decisions. That makes trust more fragile than in a conventional ad feed.

If sponsored content quietly changes an answer, users may question the entire system. If product data is inaccurate, the recommendation fails. If the highest bidder consistently outranks the best option, the assistant becomes less useful.

An AI ad network therefore needs clear rules:

• Sponsored placements must be visibly disclosed.
• Advertising should not secretly alter organic answers.
• Product claims must be supported and current.
• Sensitive conversations require stronger restrictions.
• Users need control over personalization and data.
• Merchant quality, availability, cancellation, and return outcomes should influence eligibility.

OpenAI's current advertising principles explicitly separate ads from answers. Google also labels sponsored formats and is expanding transparency for AI-generated advertising. These choices are not cosmetic. They define whether conversational commerce can scale without damaging user trust.

Why Commerce Infrastructure Becomes the Moat

AI models can interpret intent, but they do not automatically have complete commerce context.

They still need infrastructure that can:

1. Ingest and normalize merchant product data. 2. Connect publisher and creator content to products. 3. Understand inventory, pricing, promotions, and fulfillment in real time. 4. Match individual products with conversational intent. 5. Route recommendations and sponsored offers across AI, search, publisher, creator, and retail surfaces. 6. Measure purchases, cancellations, returns, margin, and repeat behavior. 7. Feed outcomes back into the next decision.

The model provides intelligence. The network provides supply, distribution, economics, and feedback.

That is why the strategic advantage may not belong only to the company with the best chatbot. It may belong to the infrastructure that connects the most useful commerce data and converts intent into successful transactions.

The Aeris View

Aeris is being built for this transition.

Our view is that commerce media and AI advertising are converging into one product-level conversion network.

Merchant catalogs create supply. Publishers and creators create context and influence. Consumers express intent through search, content, and conversation. A conversion algorithm evaluates individual products and routes them toward the channels and shoppers where they are most likely to create successful commerce outcomes.

Campaigns remain useful for budgets, objectives, and brand controls. But the intelligence beneath them must operate at the product and decision level.

The future AI ad network will not simply ask:

> Who should see this ad?

It will ask:

> What is this person trying to accomplish, which product can help, which source should validate it, which offer creates value, and what outcome should the system learn from?

That is a much larger opportunity than putting ads inside a chatbot.

It is the creation of a new commerce network built around decisions.

Sources

• Google: A new generation of ads for the AI era of Search
• Google: New tools for retailers in an agentic shopping era
• OpenAI: Ads in ChatGPT
• OpenAI: Ads in ChatGPT — The Basics
• IAB: 2026 Outlook Study