Excerpt: The sandbox demonstration shows how AI agents can move beyond product analysis to connect funding and investment in a single workflow
HONG KONG, Nov. 3, 2026 /PRNewswire/ -- EX.IO Lab, the innovation arm of Hong Kong-regulated virtual asset trading platform EX.IO, today showcased an Agentic Payment proof of concept (PoC) at Hong Kong FinTech Week. Conducted in a bank's sandbox environment, the PoC demonstrated an end-to-end, AI agent-driven workflow—from transferring funds out of a bank account to investing in wealth management products on EX.IO.
The demonstration marks a practical step towards Agentic Finance, showing how AI agents can move beyond analysing financial products to coordinating the steps needed to fund and execute an investment.
AI's role in finance is expanding beyond research and analysis into transaction workflows. Understanding a product is only part of the journey: users still need to move funds and complete an investment. EX.IO Lab's PoC addressed this gap by testing the connection between bank-held funds and investment products, bringing analysis, funding and execution into a single workflow within a sandbox environment.
From Insight to Action: Making Capital Allocation More Seamless
Investing through digital platforms often involves switching between banking and investment apps, transferring funds and placing orders in separate steps. AI may help users assess their options, but acting on that analysis can still be a fragmented process.
The PoC demonstrated how these steps could be brought together. With explicit user authorisation, defined risk limits and appropriate compliance controls, agent-driven workflows could reduce manual steps and make it easier to move funds from a bank account into a suitable investment.
The aim is not to give AI unrestricted authority over investment decisions. It is to connect analysis and execution more effectively, while keeping users in control of how their money is used and the risks they take.
The live demo showed an agentic payment workflow leading to a subscription for a real-world asset (RWA) product on EX.IO. The platform offers a range of RWA products catering to different risk appetites, return objectives and liquidity needs. That choice matters: individuals and businesses have different priorities for their capital, and no single product can meet every need.
Connecting this product range with agent-driven funding workflows could help eligible users find and access suitable investments more efficiently, subject to suitability requirements and product terms. As these capabilities develop, potential use cases include putting individuals' idle funds to work, managing short-term corporate liquidity, and managing platform cash balances between incoming and outgoing payments.
Better capital efficiency does not mean guaranteed higher returns. It means reducing unnecessary steps and idle balances so that funds can be allocated more effectively for their intended purpose. Product risk, investment horizon and redemption terms remain central to every allocation decision.
Building the Foundations for Agentic Finance
Leading payments and technology companies are already developing infrastructure for agent-led transactions. In 2025, Visa and Mastercard announced initiatives built around user authorisation, payment credentials and agent verification. Google's Agent Payments Protocol (AP2), introduced the same year, set out a framework for authorisation and payments across platforms, supporting payment methods that include instant bank transfers.
Together, these initiatives point to an emerging foundation for AI agents to transact on users' behalf—not yet to widespread adoption.
The potential commercial impact is significant. McKinsey estimates that agentic commerce could orchestrate US$3 trillion to US$5 trillion in global goods sales by 2030, assuming continued growth in AI adoption and merchant readiness. The estimate covers goods only, excluding services and B2B transactions; it is not a forecast of payment processing revenue or the size of the RWA market.
For financial services, the opportunity extends beyond how a purchase is made to how money is managed before and after it. Reducing friction between analysis, transfers and investments could make capital allocation more convenient and create new distribution channels for financial products. This is a potential growth area for Agentic Finance, rather than an established market of that scale.
Hong Kong is also advancing the infrastructure for tokenised finance. In 2025, the Hong Kong Monetary Authority (HKMA) launched EnsembleTX to move trials into real-value transactions within a controlled environment. Initial focus areas included using tokenised deposits to transact in tokenised money market funds, as well as real-time liquidity and treasury management.
At this year's Hong Kong FinTech Week, the HKMA also announced Project SAGE to support industry exploration of Agentic Finance under proportionate governance frameworks, alongside a discussion paper on banks' use of Agentic AI. These developments highlight the growing relevance of connecting funding and investment workflows.
Connecting Capabilities, Maintaining Control
Agentic Finance requires more than sophisticated analysis. It also needs clear user permissions, dependable execution and a sound understanding of product risks and user needs. Wider adoption will depend on security, interoperability and compliance with applicable regulatory requirements.
EX.IO Lab's PoC focused on one practical part of that foundation: connecting bank-held funds with investment products. By bringing product analysis, bank transfers and investment execution into an end-to-end workflow, the demonstration provides a basis for exploring more efficient ways to allocate capital—and shows how AI agents could support financial activities beyond payments.
This PoC was conducted in a sandbox environment. It does not constitute a commercial launch, regulatory approval or public availability of the services demonstrated. The personal, corporate and platform cash management use cases described above are potential applications; not all were tested in this PoC. Any future deployment will be subject to applicable local regulations, user eligibility, suitability assessments and product terms. Investments involve risk, and returns are not guaranteed.
About EX.IO
EX.IO is a regulated virtual asset trading platform (VATP) licensed by Hong Kong's Securities and Futures Commission (SFC). Licensed in December 2024, it was among the first platforms to receive a full licence after operating under Hong Kong's "deemed-to-be-licensed" regime. EX.IO is also the only VATP selected for the programme run by Hong Kong's Office for Attracting Strategic Enterprises (OASES).
EX.IO connects traditional finance (Web2) with the Web3 ecosystem, offering compliant trading, payment and settlement services alongside a range of value-added financial services.
Media enquiries: pr@ex.io
For more information, please visit: https://www.ex.io/en
References
[1] Visa Intelligent Commerce — official announcement, 30 April 2025.
[2] Mastercard Agent Pay — official announcement, 29 April 2025.
[3] Google Agent Payments Protocol — official introduction, 16 September 2025.
[4] McKinsey, The agentic commerce opportunity, 17 October 2025.
[5] HKMA announcement of EnsembleTX, the next phase of Project Ensemble, 13 November 2025.
** This press release is distributed by PR Newswire through automated distribution system, for which the client assumes full responsibility. **
EX.IO Lab Showcases Agentic Payment Proof of Concept at Hong Kong FinTech Week, Linking Bank Transfers to Investments
