How Banking teams in Hong Kong automate repetitive work with AI while respecting the PDPO and sector rules — implemented by dgm on osFoundry.
dgm is an independent osFoundry integration partner — not affiliated with osFoundry’s maker (OS LLC), and dgm has no completed client integrations yet.
Automation is where AI pays for itself in banking — but the goal is a measurable reduction in manual work on a specific workflow, not ‘AI everywhere’. Here is a sensible way to approach it in Hong Kong.
What to automate first in banking
Good first candidates are high-volume, repeatable and text- or data-heavy: real-time fraud and transaction-monitoring, AML and suspicious-activity detection and customer-service copilots over policy and product knowledge are typical. Avoid starting with one-off or highly bespoke work — the return is harder to prove.
A practical automation sequence
- Pick one repetitive banking workflow — for example real-time fraud and transaction-monitoring — and write down the current steps and time spent.
- Set a baseline so you can measure improvement, and confirm where the data lives and whether it must stay in Hong Kong.
- Build a small automation with a human in the loop, check its output against the regulator expectations that apply, then expand.
| Stage | Focus |
|---|---|
| Scope | One workflow, current steps, time spent |
| Baseline | Measurable starting point + data-residency check |
| Pilot | Human-in-the-loop build, checked against compliance |
| Expand | Roll out once value is proven |
Compliance while you automate
Banks are authorised institutions supervised by the Hong Kong Monetary Authority (HKMA), which has issued binding guidance for its sector: the 2019 Big Data Analytics and AI consumer-protection principles and, from 19 August 2024, consumer-protection guidance on generative AI in customer-facing applications. Banking is Hong Kong’s canonical high-stakes AI sector — the HKMA expects governance, fairness, transparency and data-privacy controls applied across the AI lifecycle, especially for customer-facing uses, making AI a board-level governance matter. Because there is no standalone binding AI Act in force in 2026, the constraints to design around are the PDPO (collection, use, security and the PCPD’s AI Model Framework recommendations) and the sector rules above.
Keeping automation in Hong Kong
Customer financial data and HKMA outsourcing expectations push many banks toward in-region or self-hosted deployment, even though Hong Kong has no general localisation mandate. osFoundry’s managed cloud pins data to the US, EU or Japan — it does not currently offer a Hong Kong managed region (its nearest managed region is Japan). To keep data in Hong Kong, the honest path is self-hosting osFoundry (BYO Cloud) inside a Hong Kong cloud region such as AWS Asia Pacific (Hong Kong) ap-east-1, Microsoft Azure East Asia (Hong Kong SAR) or Google Cloud asia-east2 (Hong Kong), or running models locally on-device. osFoundry can run your chosen model under one layer and be self-hosted in a Hong Kong region or run locally for sensitive workflows.
Where dgm fits
dgm is an independent integration partner that helps Hong Kong businesses adopt osFoundry — scoping a first use case, handling the build, and connecting AI to the systems you already run. dgm can build the first banking automation with you and keep a human in the loop. dgm is independent of osFoundry’s maker (OS LLC) and has no completed client integrations yet, so everything described here is a service offered, not a past result. If you want to scope a practical first project, dgm can help you map it out.