Dyna.Ai Selected for HKMA's GenAI Sandbox to Advance AI-Powered Banking

AI should be an enabler, not an end in itself
Elsie Leung, Dyna.Ai's Hong Kong head, on the company's philosophy for responsible AI deployment in banking.
Mark

So Dyna.Ai got picked for this sandbox. What does that actually mean for them?

Mimi

It means they're one of fourteen technology companies the Hong Kong Monetary Authority is letting test their AI in a real banking environment. They're not building in isolation anymore—they're working with actual banks, under regulatory oversight.

Luke

But how much of a real environment is a sandbox? The HKMA is watching. The stakes are lower than production.

Mimi

True, but that's the point. It's controlled enough to be safe, real enough to matter. They get to see if their voice AI actually works when a customer calls a bank.

Mark

What makes their voice AI different from other chatbots?

Mimi

Speed, mostly. They're responding in under two hundred milliseconds. They also handle fraud detection while they're talking to you, and they work across multiple languages and accents without falling apart.

Luke

Two hundred milliseconds—is that measured in the lab or in actual calls?

Mimi

The source doesn't specify. It's their specification.

Mark

And the company itself—how established are they?

Mimi

Founded in 2024. So they're brand new. But they've already got operations across Asia, Latin America, and Africa.

Luke

That's fast growth for a year-old company. Do we know who's funding them or who their actual customers are?

Mimi

The source doesn't say. We know they work with banks, insurance companies, and financial services firms, but no names.

Mark

So the sandbox selection is partly about proving they can scale?

Mimi

Exactly. They need to show regulators that their technology is safe and reliable at the scale banks actually need. This is their chance to do that with HKMA watching.

  • A company barely a year old is now operating inside one of Asia's most scrutinized financial regulatory environments, raising the stakes for what AI must prove to earn institutional trust.
  • The HKMA's second sandbox phase deliberately narrows its focus from possibility to accountability — 60 proposals reduced to 34 partners, with safety and reliability now the governing criteria.
  • Dyna.Ai's Voice Agents must perform in real banking conditions: sub-200ms response times, 99% accuracy across accents and noise, fraud detection running beneath every conversation.
  • The appointment of a dedicated Hong Kong head signals that Dyna.Ai is not treating this as a pilot — it is planting a regional anchor as it expands across Asia, Latin America, and Africa.
  • The sandbox now becomes both credential and crucible — a structured arena where the company must demonstrate that voice AI can handle sensitive financial data without sacrificing compliance or security.

In the layered world of financial technology, where trust is currency and caution is architecture, a one-year-old Singapore AI firm has earned a seat at one of Asia's most deliberate regulatory tables. Hong Kong's Monetary Authority, sifting through more than sixty proposals, chose Dyna.Ai as one of fourteen technology partners to test whether generative AI can serve banking's most demanding standards — not merely what is possible, but what is safe, reliable, and worthy of public confidence. The selection is less a celebration of novelty than a quiet signal that responsible innovation, when built with compliance at its core, can find its place even in the most guarded institutions.

Dyna.Ai, a Singapore-based AI company founded in 2024, has been named one of fourteen technology partners in the Hong Kong Monetary Authority's generative AI sandbox program. The announcement, made in mid-October, places the young firm alongside established institutions such as HSBC, Bank of China, and Bank of Communications in a second phase of experimentation that carries considerably higher expectations than the first.

The HKMA and its partner Cyberport narrowed the field from over sixty proposals to a final cohort of twenty banks and fourteen tech providers. The shift in emphasis is deliberate: where the program's first phase explored what AI could do, the second asks what it can be trusted to do — reliably, safely, and without introducing new vulnerabilities into banking systems.

Dyna.Ai's offering centers on Voice Agents — conversational AI systems capable of handling real-time customer interactions across English, Mandarin, Cantonese, and varied accents, with response times under two hundred milliseconds and accuracy approaching ninety-nine percent even in noisy environments. The platform also incorporates deep learning for fraud detection, identifying suspicious patterns as they surface within live interactions.

Elsie Leung, newly appointed as the company's Hong Kong head, described the selection as confirmation that AI should function as a meaningful enabler within financial services — not a novelty, but a tool built with safeguards from the ground up. Her appointment itself reflects Dyna.Ai's broader regional ambitions, with Hong Kong serving as a strategic anchor for its financial services expansion across Asia and beyond.

For a company that did not exist two years ago, entry into the HKMA sandbox is both a validation and a test — a structured opportunity to prove that voice-based AI can meet the exacting standards regulators will require of any system entrusted with customer accounts and high-stakes transactions.

Dyna.Ai, a Singapore-based artificial intelligence company founded just last year, has been selected as one of fourteen technology partners in the Hong Kong Monetary Authority's generative AI sandbox program. The selection, announced in mid-October, places the company alongside major financial institutions including Bank of China, Bank of Communications, and HSBC in a carefully curated second phase of experimentation designed to test how AI can work safely and reliably within banking systems.

The HKMA and its partner Cyberport winnowed the field from more than sixty proposals to arrive at this final cohort of twenty banks and fourteen technology providers. The shift from the program's first phase is deliberate and significant: where early exploration focused on possibility, the second phase narrows its lens to what actually works, what can be trusted, and what serves customers without introducing new risks into the financial system.

Dyna.Ai's contribution centers on what it calls Voice Agents—conversational AI systems that handle customer interactions in real time. The technology combines several layers: automatic speech recognition, text-to-speech synthesis, multi-turn dialogue management, compliance checking, and operations fast enough to respond in under two hundred milliseconds. The system works across English, Mandarin, Cantonese, and various accents, maintaining nearly ninety-nine percent accuracy even in noisy environments. Beyond conversation, the platform incorporates deep learning for fraud detection, flagging suspicious patterns as they emerge.

Elsie Leung, who recently took on the role of Hong Kong head for Dyna.Ai, framed the selection as validation of the company's approach to responsible innovation. In a statement, she emphasized that AI should function as an enabler rather than an end in itself—a tool that financial institutions deploy meaningfully and at scale, with safeguards built in from the start. The appointment of Leung itself signals the company's commitment to the region; Dyna.Ai is expanding across Asia, Latin America, and parts of Africa, but Hong Kong represents a strategic anchor for its financial services work.

The company's timing is worth noting. Founded in 2024, Dyna.Ai has moved quickly to establish itself in markets where financial regulation is sophisticated and where banks are actively seeking AI solutions that can handle sensitive customer data and high-stakes transactions. The HKMA sandbox offers a structured environment to prove that voice-based AI can work in banking without cutting corners on compliance or security. For Dyna.Ai, selection into the program is both a credential and a testing ground—a chance to demonstrate that its technology can meet the standards Hong Kong's regulators will demand of any AI system touching customer accounts.

Innovation that creates real-world value. We believe AI should be an enabler, and our focus is to help financial institutions apply it meaningfully, responsibly, and at scale.
— Elsie Leung, Head of Dyna.Ai, Hong Kong
Contact Us FAQ