A financial institution can today onboard a client on another continent in minutes, run its systems on infrastructure it has never seen, and screen that client with models it did not build. Digital onboarding, cloud computing, AI-driven compliance and global investment platforms have removed many of the geographical constraints that once shaped financial services.

Operations have become borderless. Accountability has not, and the distance between the two is where the real risks of this transformation now sit.

The direction of travel is real, and I do not dispute it. Over three decades in this industry, I have watched financial institutions move from location-based businesses, defined by where their offices and clients were, to globally connected service providers defined by infrastructure and distribution.

A trader in Tokyo, an investor in Dubai and a compliance analyst in Europe can now work on the same platform, in the same session, with the same data. Twenty years ago, connecting them required far more infrastructure and coordination. Today, much of that can sit within one well-designed stack.

Yet the conclusion that geography has stopped mattering is, in my opinion, wrong. Technology has loosened operational geography; it has not dissolved legal geography. The border no longer sits only at the firm’s address. It also sits at the client’s location and the regulator’s perimeter.

An institution’s ability to reach a client does not establish its permission to serve that client. Every market it enters brings its own rules on licensing, conduct, data and financial crime. The more borderless the operation, the more jurisdictions it touches. A borderless institution is, in practice, a multi-jurisdictional one, and that is harder to run, not easier.

The UAE shows how regulators are responding. In February 2026, the Central Bank of the UAE announced a partnership with Core42 to build sovereign financial cloud infrastructure for the UAE financial sector, including licensed financial institutions. The initiative is intended to strengthen data sovereignty and national control. That month, it also issued guidance on the responsible use of artificial intelligence and machine learning, emphasising meaningful human oversight proportionate to consumer risk.

The Capital Market Authority has meanwhile issued repeated warnings in 2026 against firms and platforms offering regulated financial services without its authorisation, including Ethr.trading, which was promoted through an Instagram account. Reach does not confer permission. My own experience confirms it: serving clients in this region legitimately meant building a licensed, locally governed presence, however global the group behind it already was.

The risks that follow are specific. A platform can drift into serving clients where it lacks the required authorisation, not through intent but through weak controls on who is being onboarded from where.

Cloud concentration and data residency can turn an efficiency decision into a supervisory problem. Remote onboarding faces synthetic identities and manipulated documents, while AI-driven screening must contend with adversaries equipped with similar tools. The most insidious risk is false confidence: the assumption that because a process is automated, it is compliant.

This is why compliance cannot be a box-ticking exercise appended after growth. In a borderless institution, it is part of the design of market access itself: which markets, under which licence, with which controls, decided before the first client is onboarded rather than after the first regulatory letter arrives.

Firms that treat it this way are better placed to scale with fewer disruptions and less reputational exposure. Firms that do not tend to discover the limits of borderlessness at the worst possible moment.

Ultimately, and this is the point most often underestimated, none of these technologies manages risk on its own. People remain accountable. AI can surface more signals than any team could review and assist with complex analysis. It cannot assume responsibility for deciding which jurisdiction’s rules govern an ambiguous case, when a model’s output should be overridden, or how to engage a regulator when something goes wrong.

The borderless institution therefore depends more, not less, on professionals with cross-market experience, regulatory judgement and the authority to act on it.

Technology decides where an institution can reach. Licensing decides where it may operate. People decide whether it should.

By Roberto d’Ambrosio, CEO of Axiory

Related post

View all