Summary

• Agentic systems can streamline campaign planning, activation, and measurement while retaining human control.
• Duffield points to the quality and governance of advertising data as critical safeguards.
• He argues that marketers’ strategic judgement and accountability remain necessary across fragmented channels.

Advertising has entered its agentic AI era.

Across the industry, the enterprise AI conversation has been framed around bigger models offering more reasoning and autonomy. Its promise is simple: give the AI system an objective and let it determine the rest. But as these systems become more capable, many assume the end goal is to remove humans from decision-making altogether.

I believe that is the wrong ambition.

The real opportunity for agentic AI is not to replace human expertise, but to amplify it. Its greatest value lies in helping marketers make better decisions, faster and with greater transparency at scale. The future of advertising will not and should not be defined by autonomous machines, but by the combination of human judgement and machine intelligence working closely together.

Human expertise must always remain central

The advertising industry has spent more than two decades automating media buying. Today, there’s not much manual work that hasn’t already been streamlined.

Our AI-driven platform, Kokai, evaluates more than 20 million advertising opportunities every second and over a trillion every day. At this scale, AI is essential. The question now is no longer whether AI belongs in advertising, but how it can create greater value. At The Trade Desk, we believe the answer lies in orchestration, not autonomy.

The role of AI is to aid marketers in navigating complexity, uncovering opportunities and acting with greater precision.

We’ve seen this principle come to life through our work with agency partners, including Stagwell, and its Koa Agents. These agentic tools streamline campaign planning, activation and measurement, without sacrificing control. This means marketers spend less time on execution and more time on strategy. Buyers define their goals, audiences and priorities, while AI helps optimise performance and surface recommendations in real time.

Human strategy remains at the centre.

Agentic systems should not decide brand strategy, determine business priorities or control budgets. Those decisions require context, experience and accountability. AI’s greatest value lies in making existing intelligence more accessible, not in replacing human judgement. And the most successful AI systems will be those that combine the speed and scale of machine intelligence with the judgement and creativity that only people can provide.

Trust can only start with quality data

The effectiveness of any agentic system depends on the quality of the information it can access.

AI does not eliminate the consequences of poor data. It magnifies them. Small gaps in data can quickly compound to become larger gaps in decision-making and outcomes when systems operate at scale.

That is why the future of agentic AI depends as much on governance and having the right guardrails as it does on models.

Advertisers need confidence in the data their AI systems use, the actions those systems can take and the transparency behind every recommendation. Agentic AI must operate within clear guardrails, grounded in trusted data and subject to human oversight.

In advertising, trust is not something that can be added later. It must be built into the system from the beginning.

The future is collaborative

As agentic AI evolves, some predict a future where every advertising channel operates through a single, fully autonomous system. The reality is likely to be far more nuanced.

Retail media, connected TV, digital audio, social and the open internet each operate with different commercial models, data signals and measurement frameworks. No AI system will erase those differences.

What AI can do is make them easier to navigate.

Agentic systems can help marketers more easily connect insights across channels and understand how different investments contribute to campaign performance – reducing complexity, without removing human control. As a result, marketers can make smarter decisions with less friction.

That is where the true value lies.

Human intelligence remains the competitive advantage

The winners of the agentic AI era will not be the companies that build the most autonomous systems.

It will be the companies that build the most trusted, transparent and effective ones.

The future belongs to organisations that use AI to scale human intelligence, not replace it. Marketers will always need accountability, creativity and strategic judgement. AI can make those qualities more powerful, but it can never substitute for them.

In advertising, the most important agent will continue to be the human behind the decision. That is why the future of agentic AI is not autonomy. It is human intelligence, amplified at scale.

About the author
Phil Duffield is VP Northern Europe at The Trade Desk. With over 20 years’ experience within the media and ad technology space, Phil leads The Trade Desk’s business and operational growth in the UK, Nordics and the Netherlands, driving market strategy, agency and brand partnerships.