Issue 13 · Nov 2–8, 2026Editions · Glossary · Archive · Method · RSS · HU / ENG

News of the AI era, curated weekly

Glossary › Agents

agentic advertising

advancedexperimental

Advertising where campaigns are planned, purchased, and refined independently by AI agents, with humans providing only the goal and framework.

In plain terms

Online advertising remains largely automated: ad spaces change hands in milliseconds via automated auctions, and machine learning increasingly fine-tunes bids. However, campaign parameters are typically set by humans, who decide whom the ad should target, where it runs, at what price and with what message. In agentic advertising, software takes over this work too: an AI agent receives the task and carries out the steps itself.

It is like not shopping yourself at the market, but trusting an experienced buyer: 'bring fresh vegetables for a ten-person dinner for 20,000 forints.' They look around, bargain, decide, and settle accounts at the end. Here, the buyer is an agent, the market is the supply of advertising spaces, and on the other side there is often also an agent, selling on behalf of the publisher or platform.

The term is also used in another related sense: for advertising aimed not at humans but at agents. If AI assistants increasingly handle purchases on your behalf, brands must persuade them, not grab your attention. In a broader sense, advertisements appearing in AI assistants' responses are sometimes included here.

How it works

  1. The advertiser states the goal in everyday language: for example, 'reach families with young children in Budapest from a monthly budget of 500,000 Hungarian forints, with the goal of online store purchases.'
  2. The buyer-side agent breaks this down: it selects target groups and channels, creates or requests ad variations, and allocates the budget.
  3. It addresses seller-side agents, i.e., publisher and platform systems, via a common language and standardized protocol: what is the offer, how much does it cost, and under what conditions.
  4. It closes deals, launches campaigns, then continuously monitors results and reallocates funds to where performance is better.
  5. Humans set limits (spending cap, prohibited content, exclusion of environments unsuitable for the brand), approve major decisions, and receive a report at the end.

An everyday example

As the owner of a small online store, you do not set up separate campaigns for every platform; you simply tell the agent who you want to reach and with what budget. As a consumer, you encounter it when your AI assistant recommends a vacuum cleaner, and one of the offers is marked as sponsored.

Who is affected

marketerInstead of manually setting up campaigns, you increasingly need to define goals, constraints, and checkpoints for the agent. → Marketing
entrepreneur, executiveYou can run a complex campaign with a smaller team, but you must decide how much spending freedom to grant the software.
developerAdvertising systems must be equipped with interfaces and protocols callable by agents. → Code
everyoneRecommendations from AI assistants may increasingly be backed by paid placements, so it is worth checking for disclosure labels.

A common misconception

MythAgentic advertising is just a new name for long-existing automated, programmatic ad buying.
In realityThe programmatic system bids and optimizes within the framework of a campaign set by humans, even with machine learning. The agent, however, plans the campaign itself, negotiates with the other party, and makes adjustments during the process to meet the given goal. It does not take the strategy and responsibility off you.

Limits and risks

Similar, but not the same

Related concepts

Broader

AI-agent, agentic AI, digital marketing

Narrower

buyer-side agent, seller-side agent, Ad Context Protocol

On the edge

programmatic advertising, agentikus kereskedelem, Model Context Protocol, multi-agent system, generative search engine optimization

A brief history

The term emerged in the advertising industry with the rise of agentic AI in 2024–2025. In October 2025, a group of ad-tech companies presented the Ad Context Protocol (AdCP) as an open standard, which provides a common language for advertising agents. Parallel to this, the industry's standards body, the IAB Tech Lab, also began working on its own agent-based standards.

Where things stand

At the end of 2026, major platforms and ad-tech companies are announcing agent-based solutions one after another, and common standards are taking shape. However, the bulk of advertising spend still goes through traditional, human-configured systems.

Status: 2026 November

In the news

Appears in 5 Agent Age stories; the curve shows weekly occurrence in the news magazine.

Also known as

agentic advertising · agentic advertising · agentic ads · agent-based advertising · agent-based advertising · agentic media buying · AI-agent advertising

Sources

This entry was machine-edited and has undergone machine expert review; the “In the news” section updates automatically from Agent Age stories.