AR in banking: real use cases, costs, and risks

If that evaluation points toward building a custom AR platform, the development effort itself carries its own cost and complexity considerations.

AR in banking at a glance

Augmented reality banking uses are still pilot-stage, not a mainstream channel yet. Industry research indicates that a growing percentage of banks are piloting or deploying AR-powered customer-facing features. Until that data is confirmed, treat "pilot-stage" as a working description rather than an established fact.

The clearest use cases in reality banking cluster around three areas: AR-powered customer onboarding, biometric authentication overlaid on a live camera feed, and interactive holographic workstations that help branch staff pull up account information faster. Two recurring friction points turn a strong pilot into a stalled build: ARKit tracking drift in low-light branches, and core banking middleware integration delays that commonly run 6-12 weeks.

Two technical paths split the field.

Marker-based AR handles kiosk and signage cases well. Markerless AR is what a real virtual branch experience inside a bank's own mobile app actually needs, and it's harder to stabilize. Each path carries distinct KYC workflow and data-handling considerations under GDPR and PSD2 SCA, so finance and compliance teams should read the fine print before committing engineering support to either route.

What is augmented reality in banking? (Marker-based vs markerless)

Augmented reality in banking overlays digital data (balances, biometric prompts, wayfinding cues) onto a live camera feed, and the tracking method you pick determines where it works.

Marker-based AR anchors that overlay to a fixed visual trigger: a QR code on a branch counter, a printed statement, an ATM decal. It is cheap to build, near-instant to calibrate, and forgiving in the mixed lighting most branch lobbies actually have.

Markerless AR, built on ARKit (iOS) or ARCore (Android), tracks surfaces and depth in real time without a printed trigger. That is what powers AR-based onboarding inside a mobile app: pointing a phone at an ID card, a physical space, or a signature field and having the app render guidance directly over it.

The tradeoff is stability. Markerless tracking drifts more in low-light branch environments and recovers slower after occlusion than marker-based setups, which is why branch kiosks often stay marker-based while customer-facing mobile experiences go markerless.

For a bank deciding between the two, the rule of thumb holds across most use cases: fixed physical locations (branches, ATMs, holographic workstations) favor marker-based AR; anything a customer carries in their pocket favors markerless AR via native SDKs. That split mirrors the broader market too, MarketsandMarkets projects the markerless segment to lead AR glass adoption through 2030, ahead of marker-based approaches.

How AR differs from VR and mixed reality in financial services

Augmented reality overlays data onto the live world. Virtual reality replaces that world entirely. Mixed reality lets digital and physical objects register together in space, so a hologram can sit on a real desk. In banking, that difference decides what a pilot can ship in one quarter versus one year.

AR runs on a customer's own phone camera, needs no headset and no new branch hardware, and slots into an existing KYC workflow because the app already handles biometric authentication and document capture. VR and MR require dedicated devices and a different budget case entirely.

The global AR market is projected to exceed $87B by 2029, up from roughly $30B in 2024, though GlobalData notes financial services adoption still lags other sectors. That gap tends to rank AR as the near-term build for most banks, with VR and MR treated as a later buy once the use case is proven.

Real-World AR banking use cases (2025)

Real-world AR banking uses in 2026 cluster into four buckets: customer onboarding, data visualization, virtual branch experiences, and wearables-based authentication. Each carries a different build timeline and a different regulatory footprint, and the scale of deployment varies just as widely.

Citi built one of the earliest bank-specific AR trading tools: a Holographic Workstation, developed with design studio 8ninths on Microsoft HoloLens, that renders market data as 3D holographic spheres and lets traders pull up volume trends and collaborate with remote clients using voice and gesture. Citi released the proof of concept in March 2016, making it the first bank to publicly demonstrate an AR trading application.

Bank of America's Erica assistant, one of the most widely used AI banking tools with over a billion client interactions, already gives customers a spending breakdown through its Spend Path feature. It's a text, voice, and tap interface today, not a camera-based AR overlay, but it shows the kind of spending-insight use case AR-based card scanning would extend if a bank chose to build it.

Westpac ran an early point-of-interest banking app that geo-tags branches and ATMs using markerless tracking instead of QR-code markers. That build choice removed the friction of asking customers to scan anything, though the app itself predates today's more interactive AR tooling and was eventually retired.

3D financial data visualization has the clearest ROI among these examples, particularly for wealth management. A relationship manager holds a tablet over a printed statement and sees a 3D allocation chart render in space, useful when visualizing risk concentration communicates faster than reading a table of numbers.

Citi's Holographic Workstation extends this idea further, projecting multi-screen trading information into desk space without extra monitors. Nearly a decade after its 2016 debut, it remains closer to a proof of concept than a trading-floor standard, but it points toward where augmented reality banking interfaces for finance professionals are headed.

WebAR removes the app-install barrier entirely.

A customer taps a link in a statement email and sees a mortgage amortization overlay directly in the browser, no App Store detour, no update cycle. WebAR tends to shorten compliance review timelines for exactly this reason: reviewers evaluate one web experience instead of a native binary queued through Apple's and Google's app stores.

Analyst estimates on future adoption vary widely and should be treated as directional rather than confirmed, not a number to plan a business case around on their own.

Wearables remain the least mature use case. Smart-glasses trials for teller-side authentication exist, but consumer-facing wearable reality banking stays experimental until battery life and social acceptability catch up.

AR-powered customer onboarding and KYC workflows

AR-powered customer onboarding replaces the multi-step KYC workflow, ID capture, selfie liveness check, proof of address, with a single guided camera session. Markerless AR overlays alignment guides on the document in real time using the phone's own camera software.

The interactive capture then hands off to biometric authentication for the liveness match, which is where most onboarding drop-off happens today. Getting this right helps banks retain applicants who would otherwise abandon a clunky, multi-app process.

The harder engineering problem sits below the SDK. The onboarding journey has to call an open banking API for account verification and a core banking middleware layer to write the KYC record, and those two systems rarely share an authentication model.

Working with Dock Financial, Netguru helped streamline KYC and compliance infrastructure, cutting manual verification steps and speeding up onboarding for new customers.

Getting this integration layer right often requires building tailored core banking infrastructure rather than forcing a one-size-fits-all middleware fit. Support from a team that has already solved these authentication mismatches once tends to shorten the second attempt considerably.

PSD2's Strong Customer Authentication requirement adds a second constraint. The biometric factor captured during AR onboarding has to bind to a registered device, not just a face template, or it fails SCA's possession-plus-inherence test.

GDPR treats that biometric template as special-category data, so storage and retention design has to happen before the AR capture flow ships, not after. Any information gathered during the scan needs a documented lawful basis in place from day one.

KYC onboarding abandonment rate averages around 10% industry-wide, and Fenergo reports that 70% of banks lose clients to slow onboarding, the gap a well-executed AR capture flow is aimed at closing.

Build-vs-buy uses come down to scope. A markerless document-capture flow bolted onto an existing mobile banking app is a lighter lift with a competent AR SDK team, though actual timelines depend heavily on the middleware work described above.

A full reality banking experience with holographic branch elements is a different regulatory and engineering commitment entirely, closer to the next section's timelines. If you're evaluating vendors, read their integration case studies closely before trusting any quoted delivery window.

Employee training and staff enablement via AR

Employee training simulation is where augmented reality earns its budget line before any customer-facing feature ships. Microsoft HoloLens lets branch staff rehearse fraud-detection scenarios, compliance escalations, and new account opening workflows against a holographic overlay of the actual teller system, rather than a static PDF or a classroom role-play.

The training use case consistently proves easier to justify than customer onboarding: no SCA constraints, no biometric data governance, no live KYC workflow to break. It's an internal build decision, not a regulatory one.

A handful of retail banks have piloted HoloLens-based onboarding simulations for new hires, layering marker-based AR onto physical branch equipment so staff practice on real hardware without real customer accounts at risk. That fits a broader enterprise pattern: industry surveys put the share of enterprises using or planning VR/AR training adoption at over 90%.

The engineering lift is smaller than a virtual branch experience build: you're rendering a fixed set of holographic workstations and scripted scenarios, not handling live financial data or biometric authentication pipelines. That makes staff-facing AR the lower-risk entry point for banks still deciding whether to build AR expertise in-house or bring in a partner for the first pilot.

Holographic workstations and 3D financial data visualization

Holographic workstations move 3D financial data visualization off the flat monitors of a trading floor and into a shared spatial workspace, where advisors and traders manipulate portfolio exposure, correlation matrices, and risk scenarios as objects they can rotate, layer, and annotate together.

The strongest use case for this class of augmented reality isn't the trading desk itself but the advisory room, where a relationship manager and a client review a risk surface side by side instead of squinting at a shared screen.

Microsoft Mesh and HoloLens-based prototypes let a wealth desk pull up a client's asset allocation as a rotating 3D model and walk through stress-test scenarios without touching a keyboard.

Gartner projects that by 2026, 75% of capital-equipment-intensive industries will use AR for cost reduction among frontline workers, a category banking's back-office and advisory desks fit into even without heavy machinery.

The build-vs-buy considerations here differ from customer-facing AR banking cases like onboarding apps. Holographic visualization tools sit closer to internal development tooling than to KYC workflow, so the compliance surface is smaller, but integration against portfolio middleware and live market data feeds runs heavier than a typical mobile use case.

We scope this as a two- to three-quarter build for a single desk, not a bank-wide overlay, given the custom data pipeline work behind each rendered position.

Virtual branches and remote advisory via AR

A virtual branch lets a customer meet an advisor rendered as an avatar, review mortgage scenarios through interactive AR overlays, and complete sign off without visiting a physical branch. That's still mostly a concept rather than a deployed pattern; what banks are actually shipping today is simpler and gets similar results. Westpac's video-based "Book a Banker" tool, for instance, drove a more than 45% increase in virtual appointments among customers in rural areas, evidence that remote advisory demand from low-footfall customers is real, even where the interface is a video call rather than a headset.

That gap matters for anyone evaluating augmented reality banking today.

Wearables, not full headsets, are the practical entry point for most finance teams. Smart glasses already support field underwriting visits and insurance-adjacent lending checks, giving loan officers instant access to policy information without switching devices. Standalone headsets stay reserved for high-value wealth advisory, where the immersion can justify the hardware cost.

The blocker is rarely the AR software itself.

It's SCA and biometric consent under PSD2 and GDPR: any face or voice capture used for authentication inside a reality banking session needs its own consent flow, separate from the KYC workflow that already gates onboarding. Read the fine print before building on top of an existing identity stack.

Over 40% of frequent branch visitors would be willing to switch to remote financial advice (McKinsey Retail Banking Survey / German Retail Banking), which helps explain why banks keep testing these uses even without full-scale rollouts yet.

Benefits of AR in banking

Augmented reality banking earns its budget line through practical uses that help banks serve customers more efficiently: fewer branch visits, faster onboarding, and tighter fraud control. A virtual branch experience can lower the cost of serving low-footfall locations, since advisory sessions run over headsets instead of physical real estate.

Marker-based AR overlays on shared documents also shorten mortgage and loan walkthroughs, letting advisors share account information in real time rather than scheduling multiple in-person meetings. Institutions piloting this reality banking model report shorter meeting cycles, though exact figures on branch-visit reduction and fraud exposure still vary by use case and deserve independent validation.

Onboarding is where the return on investment case is clearest.

AR-powered customer onboarding software paired with biometric authentication compresses document capture, liveness checks, and KYC workflow steps into a single guided session. This matters directly for PSD2 SCA compliance, since the same biometric layer satisfies the regulator's authentication requirement.

According to Gartner's immersive technology forecast for financial services, a growing share of banks are piloting AR and VR features to support customer engagement and staff training through 2027.

Pilots clear technical review faster when they attach to an existing onboarding or advisory flow rather than launching as a standalone, interactive app.

The pattern holds across other industries too: Netguru's ARrange, a retail AR furniture-visualization demo, shipped in five weeks with full feature scope, evidence that fast AR development timelines are achievable when the scope is well-defined, even if a banking pilot's compliance layer adds time a retail demo never has to account for.

AR, biometric authentication, and data privacy risk

Biometric authentication layered onto augmented reality features raises the compliance bar for banking, because facial or iris capture through an AR camera falls under GDPR's special-category data rules and PSD2's Strong Customer Authentication mandate. Any bank building AR-powered customer onboarding has to treat biometric templates as data requiring explicit consent, on-device storage, and a documented deletion path, not just another KYC workflow field.

The recurring technical obstacle is not the compliance framework itself, it is capture reliability. ARKit and ARCore tracking drifts in low-light branch environments, and a failed facial scan during SCA gets logged as a false rejection, which regulators scrutinize as closely as false acceptances.

That reliability gap is exactly why build-versus-buy matters early. A vendor SDK with proven liveness detection cuts months off a pilot compared with training a custom model on a bank's own low-light branch footage. Signius reports that 77% of European banks cite poor UX and process delays, and 45% cite complex procedures, as their top digital onboarding dropout drivers, the exact failure mode a reliable capture flow is meant to fix.

Data minimization is the practical fix: process biometric matching on-device, transmit only a pass/fail token to core banking middleware, and never let raw facial or iris data touch a shared server. It satisfies GDPR's storage-limitation principle and keeps the audit trail for PSD2 SCA exemptions simple to produce.

Key considerations and cost before building AR in banking

Total cost of ownership for Augmented Reality in banking splits into three buckets: integration with existing core banking middleware, AR SDK licensing, and compliance tooling for biometric authentication. The first bucket, not the AR layer itself, tends to drive most budget overruns on banking pilots.

The integration timeline runs longer than teams expect going in. A markerless AR onboarding flow that calls an open banking API for account verification typically needs 8-14 weeks: two to three weeks for SDK selection and proof of concept, four to six weeks wiring the app to core banking and KYC workflow endpoints, and two to four weeks for SCA and biometric compliance review before rollout.

Get a specific budget range from your own vendor conversations rather than a single industry-wide figure, since core banking stack complexity swings the number more than the AR layer itself does.

Build versus buy is the next decision. A white-label AR SDK compresses the build timeline to weeks, but locks the bank into one vendor's tracking engine, a problem if marker-based AR in branch signage later needs to share a codebase with markerless AR for mobile onboarding.

Building in-house costs more up front but keeps the open banking API integration and biometric compliance logic under direct control, which matters once GDPR and PSD2 SCA audits start asking for source-level evidence rather than a vendor's compliance sheet.

Before signing off on any AR banking case, run a two-week technical spike against your actual core banking sandbox, not a vendor demo environment. That is where tracking drift, latency, and integration gaps in existing systems surface first, well before the compliance review does.

FAQ: AR applications in financial services