For whom does the agent toll?

As AI agents begin to take on more responsibility in commerce, a quiet shift is happening. Decisions that were once made by people are now being made on their behalf.

The consumer sets the intent. The agent executes.

But this raises a deeper question. When an agent makes a decision, who is it really working for?

Agents do not just assist. They act.

Agentic systems are designed to move beyond recommendation. They search, compare, select, and in some cases, transact.

This changes the role of technology in commerce. It is no longer just influencing decisions. It is making them.

According to World Economic Forum, AI agents are expected to play an increasing role in how products are discovered and chosen, shifting power away from traditional marketing channels toward system-level decision-making.

Alignment is not guaranteed

In theory, an agent should act in the best interest of the user. It should optimise for their preferences, constraints, and goals.

In practice, it is more complicated.

Agents are built by companies, trained on data, and shaped by incentives. These incentives may include partnerships, platform priorities, or commercial relationships that are not always visible to the end user.

Research from OECD  highlights the importance of transparency and accountability in AI systems, particularly as they take on more decision-making responsibility.

If those elements are unclear, the question of who the agent serves becomes harder to answer.

The shift from visibility to selection

In traditional digital commerce, brands compete for attention. Visibility is the goal.

In agentic commerce, the dynamic changes. Products are not just seen. They are selected.

This means that being known is no longer enough. Products need to be understood by the systems making decisions. Structured data, clear signals, and machine-readable context begin to matter more than brand storytelling alone.

The agent does not browse. It evaluates.

A more useful question

“For whom does the agent toll?” is not just a philosophical question. It is a practical one.

Does the agent serve the user? The platform? The merchant? Or a mix of all three?

The answer depends on how the system is designed, what data it uses, and what incentives guide it.

That is where businesses need to pay attention.

If you are trying to understand how agentic commerce will affect your visibility and sales, Hui Newnham helps companies prepare for this shift in a practical way. You can book a free strategy call with him, and follow him on LinkedIn for clear, grounded insights.