AI in accounts payable: A threat to jobs or a catalyst for value?

Artificial Intelligence (AI) is rapidly reshaping accounts payable (AP). The question now facing finance leaders is straightforward: if AI can do the work, will it replace people?

The answer is more nuanced than many expect. AI will replace some tasks, but it will not replace the AP function. In many organisations, it will also highlight whether AP was ever set up to create value beyond processing.

The gap AI cannot close

At the centre of this debate is a simple but important distinction. Accounting is deterministic. Outputs must be correct, auditable, and compliant. There is no room for approximation.

AI works differently; it produces outcomes based on probability rather than certainty. Whether it is extracting invoice data, matching transactions or flagging anomalies, it is always making a judgment call based on patterns in data.

Even with high levels of accuracy, a gap remains. At scale, small error rates turn into a steady stream of exceptions. Those exceptions still need to be reviewed, interpreted, and signed off by someone who understands the context and takes ownership of the outcome. This is not a temporary limitation. This is how these systems work.

As a result, automation does not remove the need for AP professionals. It changes where they spend their time and where their value lies.

A familiar shift, moving faster

Finance has been here before, with previous waves of technology innovations removing large volumes of manual work. Finance tasks that once required teams of people were reduced due to systems and software. The AP function did not disappear - it evolved.

AI is the next step in that process, but it is happening faster and more visibly. Data entry, approval chasing, and routine reconciliation are being automated at pace. What remains is the part of the job that requires judgment and ownership.

The risk is not that jobs disappear entirely. It is that organisations fail to redefine the role once the processing layer is gone.

From volume to judgement

As automation increases, the way in which performance is measured will start to shift. It is no longer about how many invoices are processed but the quality of decisions being made. Working capital becomes more active, enabling better case flow, higher efficiency, more responsive operations and improved returns.  Instead of payments just happening automatically, the business is actively managing them to save money and control cash better.

Supplier risk becomes more complex, as although AI can highlight anomalies, it cannot explain them. Understanding whether something is a fraud signal or a legitimate business change still relies on human judgment. The focus of the role shifts to exceptions. Disputes, partial deliveries, and contract interpretation do not disappear – they are now a core part of the work.

Governance also becomes more important, as systems now take on more responsibility. A finance professional will define the rules, review outcomes, and take ownership when something goes wrong. In this environment, AP is no longer just a transactional function; it becomes part of how organisations manage control and risk.

The real risk is underutilisation

The bigger risk for CFOs is not that AI replaces AP teams - it is about assuming that the organisation has done enough by just being more efficient. Automation makes it easy to reduce cost – which some believe this equates to reducing headcount and processes can be accelerated. But if the role itself is not redefined, the function becomes more efficient without becoming more effective.

There is also a control risk. As automated systems operate at scale, any gaps in oversight become more visible. So, AI does not remove risk; it shifts where that risk sits. Without the right level of human involvement in exceptions and oversight, that risk becomes harder to manage.

What happens to jobs?

Some roles will change significantly. Positions focused on manual data entry, routine matching, and administrative follow-up are already becoming less viable on their own.

That is a real shift, and it should not be ignored. At the same time, the core skills of strong AP professionals remain relevant. Attention to detail, understanding of controls, the ability to spot anomalies, and knowledge of supplier relationships are still critical.

Those skills are not being replaced; they're just being applied in different ways and in more valuable contexts.

The role moves upwards, away from execution and towards oversight and decision-making.

A more valuable function, if organisations choose it

AI is not replacing accounts payable. Like other manual roles, it is removing the manual layer that has defined it for years. What comes next depends on how organisations respond.

Some will use automation to do the same work with fewer people, treating AP as a leaner cost centre. When used more strategically, others will use it to reposition the function, focusing on cash flow, supplier risk, and financial control. The difference between those outcomes is not technology; it is how the function is designed and led.

For CFOs, the question is no longer whether AI will change AP, as this has already happened. The real question that they need to explore is whether it will be used to reduce the function or to make it more valuable.