The return-to-office debate is still dividing the tech industry, but payments have an interesting contradiction. Infrastructure no longer depends on physical proximity, yet the people operating it are still expected to.

Security and access to sensitive systems have traditionally made remote work feel harder to justify in payments than in other industries. Thanks to advances in cryptography, permissioning, and cloud technologies, physical proximity is no longer a necessity for getting work done – even in payments, one of the world’s most trust-dependent industries. Monitoring data coming in or out of the systems remotely is now almost as easy as doing so from the same location.

So, if the infrastructure powering modern payments can operate virtually, can the people building and securing that infrastructure do the same?

Opportunity isn’t office-bound

Treating the office as the ultimate enabler of culture, productivity and collaboration puts a lot of faith in something that can also constrain how a company is built. It limits the pool of people you can hire from and ties businesses to highly competitive talent markets.

As education and training for technology roles improve across more markets, companies can increasingly hire based on capability rather than location. A broader talent pool also brings more variety in how people think and work, rather than encouraging companies to hire people who happen to live in the same places or come from similar backgrounds.

That flexibility still needs clear responsibilities and deadlines, but those can coexist with greater discretion over how people work. This requires trust, but, as a matter of fact, you should be working with people you can trust in the first place. In payments, where trust is fundamental to the business itself, that should be a basic expectation.

The perimeter is no longer a place

Contrary to the assumption that remote work uniformly increases the risk to financial data, it actually redistributes the risk. Some threats associated with a physical office, such as tailgating and rogue network ports become redundant. Others become more difficult to control – for instance, verifying a new hire you may not meet in person while onboarding or dealing with helpdesk social engineering when established trust channels can be synthesized.

Resorting to policy might seem reflexive, but the strongest control is scope. A home office isn’t a threat to cardholder data when no cardholder data touches an endpoint. It can remain within single-tenant cloud environments and be reached through brokered, ephemeral, and fully logged access rather than standing credentials.

Physical security is therefore a question of architecture rather than location, where the latter is no longer a security boundary but remains a legal one. Data residency, sanctions, and requirements such as DORA can still determine where an employee is permitted to work from.

When it comes to access, the same principle applies. What matters is who is accessing the system, what device they’re using, and what they need access to. It should be granted on the basis of a verified identity and an attested managed device, alongside a defined role and a clear business need. Network position is no longer the deciding factor, nor was it ever a strong proxy for trust.

Compliance then becomes much less of a separate exercise. Automated controls generate their own evidence, without anyone having to retrospectively explain ‘how we do things around here’.

But what about synthetic employee fraud?

Synthetic employee fraud is one area where fully remote working does introduce a specific challenge simply because you may not meet someone in person during hiring or onboarding. There is less opportunity to establish that the person behind the identity is who they claim to be. But trying to get better at spotting deepfakes is not an enduring solution because AI will keep moving faster than detection can tag along.

The better approach is to build defense in depth. In other words, build enough checks into the process that getting past one of them doesn’t get someone too far. Consider strong ID and liveness checks, managed devices shipped to verified addresses, and access tied to those devices – and the same standards applied to contractors and third parties.

But no identity verification is going to be perfect, so the controls have to account for failure. That’s why no single identity should carry enough access to become a serious problem on its own or give an attacker free rein. Least privilege and segregation of duties limit what an account can reach, while strong authentication, access reviews, and behavioral monitoring help catch anything that starts to look out of place. Then you have audit logs to ensure there is a clear trail for everything.

Moreover, being fully remote doesn’t necessarily mean you’ll never meet the team in person. The occasional in-person meeting still has its place, even if it’s just once a year or so to bring everyone together and maintain those relationships. A synthetic identity sure would struggle to make it that far!

We've featured the best credit card processing for small business.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit