Payments firms need to be preparing for the growing use of agentic AI agents and the new security and compliance risks posed by a world where consumers pass off their spending to trusted digital companions.
In 2026, it is extremely rare for members of the general public to trust AI to carry out transactions on their behalf, but trends in the overall use of AI indicate that this future is unlikely to be far away.
Research by UpCity suggests that half of U.S. citizens use AI voice assistants every day to look up basic info or carry out simple automated tasks on their smart devices, for example.
And in the workplace, office-based staff are becoming more and more familiar with relying on AI agents to support their day-to-day productivity.
A McKinsey investigation found that 71 percent of businesses have already deeply integrated AI into at least one core business function.
In the home, AI agents already operate in limited, controlled environments. These include help with scheduling, budgeting and basic financial tasks.
While concerns among the public over the ways in which AI spreads misinformation and its impact on the workforce persist, the use of AI agents to support everyday activities delivers on some of the early promise of artificial intelligence - that it would allow people to gain free time by handing off some of the admin of real life to capable digital helpers.
With AI tech improving at a rapid rate, the tipping point is likely to come when trust reaches a critical mass.
As the use of AI becomes more widespread and the reliability of the technology advances, payments experts fully expect the use of agentic AI agents in handling day-to-day payments.
Whether it’s paying utility bills or buying more paperclips, there are countless everyday tasks that will likely be handled automatically in the future.
If you’re a payments provider, that changes the calculus for how you think about risk and fraud more generally.
Trust goes both ways, and while consumer trust will be key to growing the use of agentic AI in payments, having confidence in a payee's identity is key to payments industry protections.
When the buyer is a human being that entails one set of controls: KYC procedures are by now well established and highly successful and catching when someone is pretending to be someone they are not.
But AIs do not necessarily behave in the same ways as normal consumers and may trigger additional checks and payment blocks just by behaving entirely normally and carrying out the tasks assigned to them without deviation.
That is giving rise to the concept of Know Your Agent (KYA) and payments firms are being advised to build understanding and processes for this new strand of anti-fraud sooner rather than later.
Providers that delay in internalising KYA face presenting consumers and businesses with unnecessarily delays and user flows filled with unwelcome friction.
In short, exactly the sorts of problems that will see clients turn to competitors.
KYA will entail hitting an entirely different set of objectives compared to KYC.
Instead of establishing human identity, systems need to ascertain if the agent making the payment is correctly linked back to a real human who is the true owner of the payment method.
However they will also need to perform checks that establish that the AI agent does in fact have the authorisation of the individual to carry out the payment, and that it is doing as it was asked correctly.
These are entirely new challenges and ask questions that the existing AML and KYC frameworks are poorly equipped to answer.
Existing KYC processes rely heavily on historical data, behavioural tracking and digital footprints, all of which do not necessarily line up well with AI agent behaviour or include the data needed to establish identity and purpose.
The need for strong KYA is even more critical than with standard human-based payments flows.
That’s because, as is already becoming clear, the power of AI gives bad actors the ability to exponentially ramp up their illicit activities.
With the power of smart AI systems at their backs, fraudsters can execute new kinds of scaled attacks linked to credential stuffing, bot attacks and identity probing.
The growth of synthetic identity fraud is also expected over the next few years, as criminals use AI to build a convincing model of another person’s identity and then further obfuscate that forgery by using an AI agent to enact the payment.
In fact, this layer of abstraction is a major issue that will need to be surmounted by compliance pros over the next few years.
With the all-important human now a layer away from the actual transaction, it becomes all the harder for payments providers to accurately ascertain if they are indeed linked to the payment method being used to carry out the transaction.
All this, and more, is why smart payments companies are bolstering their compliance and anti-fraud teams with the expertise they need now to ensure they are ready to lead the pack when agentic AI agents truly enter the marketplace.
PaymentsGenes has unique access to the hiring market and an expertise in the field that only comes from being a specialized workforce partner.
Reach out now if you’d like to understand more about how to future-proof your payments business and place yourself one step ahead of the AI revolution.
As competition increases and product offerings inevitably converge, the key differentiator for betting companies is customer experience. The way an operator handles its payments infrastructure is perhaps the most powerful lever they have to pull in this arena.
.jpg)
