- Is it really worth relying on AI agents for your company's paid campaigns?
- What is really happening in major international agencies?
- Who decides what: how does the split between AI and human work in campaigns?
- What changes in practice for an SME or an Italian company working with a media agency?
- What governance tools and approaches exist for AI agents in campaigns?
- What legal rules do you need to know before using AI agents in campaigns?
- Why everyone's talking about it now: is the market really ready or is it just hype?
- How to get started concretely: the steps to evaluate and govern AI agents in your campaigns?
- What are the most common mistakes made by those who rely on AI agents without the right rules?
- How to understand if the governance of AI agents in your campaign is working?
- So, is it worth it for your company to rely on AI agents in campaigns?
- How we see it at SHM
- Frequently Asked Questions
- Major international agencies are already using AI agents to plan, buy, and optimize paid campaigns: the change is happening now.
- The main risk is not AI 'going rogue,' but the lack of approval thresholds, human controls, and clear rules on who decides what.
- 78% of European organizations have already appointed a formal head for AI governance: if your agency hasn't, that's a question to ask immediately.
- Asking your agency how it monitors and governs AI agents in campaigns is the first concrete step to protect your budget and stay in control.
Two-thirds of media buyers are already testing AI agents for campaigns, but only 30% have scaled automation across the entire cycle. The gap between those who start and those who truly govern is still huge — and the budget at risk is yours.
Is it really worth relying on AI agents for your company's paid campaigns?
It depends on how much your agency has structured its controls: agentive AI brings speed and scale, but without clear governance, the risk of runaway spending is real.
A AI agent — meaning a software system that makes decisions autonomously, without a human approving every single action — is becoming the standard tool for buying online advertising space. It's not science fiction: according to Digiday , in just one week of October 2026, three major international groups announced active agentive systems on their clients' campaigns.
For an Italian marketing manager or entrepreneur, the right question is not "is AI ready?" but "does my agency have the rules to use it safely?".
| Your situation | Is it worth relying on AI agents? | why |
|---|---|---|
| Agency with documented approval processes and a dedicated AI point person | Yes | You can benefit from speed and optimization with low risk |
| Agency using AI but without spending thresholds or automatic alerts | Evaluate first | You need a written agreement on who approves high-impact decisions |
| Agency that cannot answer questions about AI governance | No, for now | The risk of unauthorized spending or undetected errors is too high |
| Campaign budget under €5,000/month | Wait | Governance overhead often outweighs the benefit at these volumes |
What is really happening in major international agencies?
In just a few weeks, three of the world's leading media groups have launched operational agentive systems: these are not future announcements, they are tools already used on real campaigns.
Dentsu has announced that its agency Carat is collaborating with NBC Universal, Freewheel, and the AI company Newton to launch an agent dedicated to planning, purchasing, and optimizing advertising space. A client in the luxury retail sector is already using it.
The independent agency Tinuiti has opened its proprietary measurement system directly to AI agents, bringing audience, creative, media, and measurement management under a single server. Stagwell has relaunched its media agency with an agent-based system called Machine OS, designed to react in real-time to market changes.
These are not lab experiments. These are systems in production, with real budgets from real clients. If you work with a media agency — or are considering doing so — this changes the questions you need to ask.
| Group / Agency | Agentic system | What it does independently |
|---|---|---|
| Dentsu / Carat | Agent with Newton + Freewheel | Upfront planning, buying, and optimization |
| Tinuiti | Bliss Point MCP Server | Unified management of audience, creatives, media, and measurement |
| Stagwell / Stagwell Media | Machine OS | Real-time reaction to market changes |
| WPP Media | Internal system with structured governance | Inventory analysis, AI recommendations with human approval |
Who decides what: how does the split between AI and human work in campaigns?
The model that the most advanced agencies are adopting clearly separates automatable decisions from those that require human approval, especially where there is financial impact.
The recurring keyword across all agencies interviewed by Digiday is «human in the loop» — meaning a human who remains in the decision-making loop, especially for high-impact choices. It's not an abstract ethical choice: it's a practical necessity, because AI agents make mistakes, and without checks, those mistakes translate into unauthorized spending.
- 1 · AI agent Analysis of available inventory The agent scans available ad spaces across multiple platforms and evaluates them based on campaign objectives.
- 2 · AI Agent Purchase recommendation It proposes an optimized media plan for audience, format, and timing. It does not execute yet.
- 3 · Human point of contact Approval or modification A senior media buyer reviews the proposal, checks consistency with the strategy, and approves within predefined thresholds.
- 4 · AI Agent Real-time execution and optimization After approval, the agent executes and optimizes independently, but within the approved parameters.
- 5 · Human point of contact Monitoring and alerts The team receives automatic alerts if the agent exceeds spending or performance thresholds outside the range.
What changes in practice for an SME or an Italian company working with a media agency?
The main risk is not the AI 'going crazy': it's not knowing what decisions it's making autonomously and not having written agreements on who is responsible for errors.
Think of an Italian manufacturing company that invests 20,000 euros per month in Google and Meta campaigns. Today, that expense is managed by an account manager who approves every significant change. With an AI agent, the same budget can be optimized in real-time — but if the agent decides to move 8,000 euros to a wrong channel without anyone noticing, the damage is already done.
The question to ask your agency isn't 'do you use AI?' but 'what decisions can AI make autonomously on my budget, and which ones require my approval?'.
| Italian company type | Main risk | Question to ask the agency |
|---|---|---|
| E-commerce with seasonal campaigns | Accelerated spending on the wrong products during peaks | What daily spending thresholds trigger a human alert? |
| Professional firm (B2B) | Low-quality leads optimized by AI for volume, not value | How does the agent distinguish a qualified lead from an unqualified one? |
| Manufacturing company with export operations | Incorrect geo-targeting on non-priority markets | Who approves targeting changes greater than X%? |
| Physical store network | Budget shifted to digital channels inconsistent with the region | Can the agent modify the geographical distribution of the budget? |
What governance tools and approaches exist for AI agents in campaigns?
The most structured agencies use three levels of control: automatic approval thresholds, separation between recommendation and execution, and periodic audits of the agent's behavior.
There is no universal standard yet, but the model emerging from the statements of international agencies — and that you can also request from yours — is structured on three levels. The topic is also explored in the article on how to stay in control when Google and Meta automate campaigns , which tackles the problem from the platforms' side.
| Control level | What it includes | Who manages it |
|---|---|---|
| Approval thresholds | The agent cannot spend over X€/day or modify targeting beyond Y% without approval | Account manager + client |
| Separation of recommendation and execution | AI proposes, human approves before execution on high-impact decisions | Senior media buyer |
| Periodic behavior audit | Weekly or monthly review of decisions made by the agent and results | Internal team + client |
| Automatic alerts | Real-time notifications if the agent exceeds predefined parameters | Automatic system → human contact person |
| Separation of probabilistic AI / deterministic execution | AI recommendations are marked as such, distinct from certain executions | Agentic system architecture |
What legal rules do you need to know before using AI agents in campaigns?
From August 2, 2026, the European AI Act will impose transparency obligations on AI-generated content: those who do not comply risk sanctions of up to 15 million euros.
From August 2, 2026, the obligations of Article 50 of the EU Regulation on Artificial Intelligence (AI Act) will be fully operational. In practice: if an AI agent generates texts, images, or videos used in campaigns, these contents must be recognizable as artificial. Violation is punishable by fines of up to 15 million euros or 3% of annual global turnover. The details are clearly explained by Cravero Studio in the guide to Article 50 of the AI Act .
For an Italian company entrusting campaigns to an agency, this means one concrete thing: you need to know if the creative content produced by the AI agent is correctly labeled. If it's not, the responsibility can also fall on the advertiser, not just the agency.
Why everyone's talking about it now: is the market really ready or is it just hype?
The numbers show that adoption is real but uneven: large organizations are much further ahead than SMEs, and governance is still the sector's weak point.
The IAB Europe report «Impact of AI on Digital Advertising 2026» (September 2026, 50 European respondents) shows that 86% of organizations already use AI for marketing and 72% have increased investments compared to 2025. But the current performance of AI agents in advertising operations gets an average rating of 2.76 out of 5: we are still in the maturation phase, not full confidence.
The gap between large and small organizations is clear: among companies with more than 500 employees, 86% have agentive systems at least at the second level of autonomy. Among those with fewer than 500 employees, the share drops to 48%, according to IAB Europe . For Italian SMEs, this means the risk of relying on immature tools is higher, not lower.
The article on agentic buying and automated media buying it explores market context and current adoption dynamics.
How to get started concretely: the steps to evaluate and govern AI agents in your campaigns?
Before activating any agent system, an audit of your agency's governance and a written agreement on who decides what are needed.
The starting point isn't choosing an AI tool: it's understanding how your agency already handles automated decisions. Many agencies have been using automations for years (automatic bids on Google Ads, automatic optimization on Meta) without ever formalizing them in a governance document. Next-generation AI agents amplify this problem because they act on multiple variables simultaneously and faster.
To learn more about how AI agents can scale campaigns without increasing the team, also read AI agents for campaigns: scaling without hiring .
- 1 · You (client) Ask your agency for an AI governance document It must specify which decisions the agent makes autonomously, which require approval, and who the responsible human contact is.
- 2 · You + agency Set spending thresholds and alerts Establish in writing the daily and weekly thresholds beyond which the agent cannot act without explicit approval.
- 3 · Agency Check compliance with the AI Act All AI-generated creative content must be labeled as such. Ask for written confirmation that the agency complies with Article 50 of EU Regulation 2024/1689.
- 4 · You + agency Set up a monthly report on agent behavior Performance reports are not enough: you want to know how many decisions the agent made, on which variables, and with what results.
- 5 · You (client) Review the agreement every quarter The AI agent market is moving fast. Governance thresholds and rules should be updated at least every three months.
What are the most common mistakes made by those who rely on AI agents without the right rules?
The most frequent mistake is delegating high-financial-impact decisions to AI without defining who is responsible if something goes wrong.
The original source states it clearly: 'not every decision should be delegated to an agent.' The problem is that many companies — and some agencies — have not yet drawn the line between what AI can do autonomously and what requires human supervision. Here are the most frequent mistakes we see in the market.
A related topic is unreviewed AI content: risks, rules, and how to protect yourself it is a useful read to understand where legal responsibilities are concentrated.
| Error | Concrete consequence | How to avoid it |
|---|---|---|
| No automatic spending thresholds | The agent can burn the monthly budget in a few days on the wrong audience | Define daily and weekly caps in writing |
| Confusing AI recommendations with definitive decisions | Actions are executed based on probabilities, not certain data | Ask the agency to separate probabilistic outputs from deterministic executions |
| No human point of contact appointed | If something goes wrong, nobody knows who is responsible | Include the name of the responsible senior buyer in the contract |
| Audits only on results, not on agent behavior | You can tell the campaign underperformed, but it's unclear why | Request reports on the agent's decision-making process, not just on KPIs |
| Unlabeled AI content | Risk of AI Act fines up to 15 million euros | Verify compliance with Article 50 of EU Regulation 2024/1689 |
How to understand if the governance of AI agents in your campaign is working?
Governance works when you can answer three questions: what decisions did the agent make this week, who approved them, and what results did they produce.
Monitoring the performance of a campaign managed by AI agents requires different indicators than a traditional campaign. It's not enough to look at cost per click or return on ad spend: you also need to know how the agent achieved those results. Without this transparency, you can't distinguish good performance from lucky performance.
For those managing campaigns across multiple platforms, cross-platform measurement has become a priority for 72% of media buyers in 2026 (up from 64% in 2025), according to the IAB 2026 Outlook Study : a sign that the market is realizing that agentive AI requires new measurement tools.
| Indicator | What it measures | Recommended frequency |
|---|---|---|
| Number of autonomous decisions made by the agent | How many actions were executed without human approval | Weekly |
| Human intervention rate | % of decisions blocked or modified by the team before execution | Weekly |
| Deviation from the approved budget | Difference between planned spending and actual agent spending | Daily |
| Lead quality / conversions | Does the agent optimize for volume or value? | Monthly |
| AI Act compliance | Is AI-generated content labeled correctly? | Every campaign |
| Human contact person satisfaction | Does the senior buyer consider the agent's recommendations reliable? | Monthly |
So, is it worth it for your company to rely on AI agents in campaigns?
It is worth it if your agency has an AI governance document, a designated human contact, defined spending thresholds, and complies with the AI Act. In that case, the speed and scale of AI agents are a real advantage.
A middle ground your agency already uses automations (automatic offers, Meta optimization) but has not yet formalized the rules for next-generation agents. It's worth requesting a dedicated meeting before expanding automation.
Not yet if your agency can't answer questions about who approves agent decisions or has never heard of the AI Act. In that case, the risk outweighs the benefit: it's better to wait or change interlocutor.
How we see it at SHM
For us, AI agent governance isn't a technical issue: it's a relationship issue. An AI agent managing your ad budget is a representative — and like any representative, it must have clear instructions, defined limits, and someone accountable for its actions. We believe that simplifying digital also means making it understandable what the AI does with your budget, not hiding it behind a performance dashboard.
We work with Italian companies that want to integrate AI into paid campaigns without losing control. Our approach always starts with an audit of the existing governance, then we define together the thresholds, approval processes, and monitoring indicators. If you are considering how to structure this path, the starting point is our service of strategic AI integration . For those who want to understand how AI is changing the organic visibility of campaigns, the service of SEO for AI is the natural complement. And for those managing Google Ads campaigns and wanting to understand where automation helps and where it risks getting out of hand, our team of google ads campaigns can make a concrete assessment.
You can find all articles on paid media, automation, and AI in our hub Marketing and ADV performance .
Let's consider a manufacturing company that asked us to review their relationship with their media agency after noticing unexplained discrepancies between the approved budget and the actual spending. It wasn't a fraud issue: it was simply that no one had ever defined which decisions automation could make independently. We helped structure a simple governance document — two pages — that solved the problem and improved trust in the relationship with the agency. A more sophisticated AI system wasn't needed: clarity on who decides what was.
Frequently Asked Questions
An AI agent in media buying is a software system that makes autonomous decisions: it analyzes available ad spaces, proposes or executes purchases, and optimizes campaigns without a human approving every single action. The most advanced agencies are already using them on real campaigns, with control systems that limit the agent's autonomy on high-financial-impact decisions.
The direct question is the simplest approach: ask your agency if they use advanced automation systems or AI agents on your campaigns, and what decisions these systems can make autonomously. If the agency cannot answer or does not have a governance document, it's a red flag.
The main risk is unauthorized spending: an agent without control thresholds can shift budget to wrong channels or audiences before anyone notices. Added to this is the legal risk: if creative content generated by AI is not labeled as such, the company may incur penalties provided for by the European AI Act.
Yes. From August 2, 2026, Article 50 of EU Regulation 2024/1689 requires that AI-generated content — text, images, videos — must be recognizable as artificial. This also applies to advertising creatives produced by AI agents. Violations are punishable by fines of up to 15 million euros or 3% of global turnover.
According to the IAB Europe report of September 2026, 58% of European operators expect the purchase of media space through AI agents to reach regular operational use within a year. However, the current performance of AI agents in advertising operations gets an average rating of 2.76 out of 5: we are still in a maturation phase.
There are four essential questions: what decisions does the agent make autonomously on my budget? Who is the responsible human contact? What spending thresholds trigger an alert or require my approval? Are AI-generated content labeled in compliance with the AI Act? If the agency doesn't have clear answers to all four, it's necessary to formalize the rules before proceeding.
The real question isn't whether to trust AI, but whether you've put in writing who is responsible when AI makes a mistake.
Related articles
Discover more articles exploring similar topics, selected to offer you a more complete and stimulating perspective. Each piece of content is carefully chosen to enrich your experience.
