1. Why is HSBC opening an AI research institute in Hong Kong?
HSBC will set up a global AI research institute in Hong Kong, with operations expected to begin in early 2027. David Liao, the bank's co-chief executive for Asia and the Middle East, said the institute will start with three areas of work: practical applications of artificial intelligence, post-quantum cryptography and blockchain technology. The bank also confirmed that its Hong Kong AI team, now about 150 specialists, is expected to double in size over the next two to three years, reaching roughly 300 people.
The plan turns the city into more than a regional office for one of the world's largest banks. A research institute implies that new methods, tools and prototypes will be developed there, rather than simply deployed from elsewhere. For a lender that has long described Hong Kong as central to its history and its earnings, that is a meaningful statement about where it wants its technical expertise to sit.
The announcement also carries a message for the city's regulators. It arrived against the backdrop of a separate HSBC decision to build an AI centre of excellence in Singapore, a move that drew questions from the Hong Kong Monetary Authority. Taken together, the two projects show a bank trying to grow its artificial intelligence capability in more than one financial hub while keeping its home market at the center of that effort.
2. What will the institute work on first?
The initial remit has three parts. The first is AI application cases, meaning the use of machine learning and generative models for tasks inside a bank, such as handling customer requests, supporting staff who write and review documents, and spotting unusual patterns in transactions. Applied work of this kind is the most visible and the most immediately useful, which is why it sits at the head of the list.
The second is post-quantum cryptography. This is a branch of security research that designs encryption able to withstand attacks from future quantum computers. Today's widely used public-key systems rely on mathematical problems that a sufficiently powerful quantum machine could, in principle, solve far faster than any conventional computer. No such machine exists at the necessary scale today, but banks hold data that must stay confidential for many years, so they have an incentive to prepare early.
The third is blockchain technology. HSBC has not linked the institute to any specific token or network, and the announcement should be read as a statement about research priorities rather than a product launch. Even so, naming blockchain alongside AI and quantum-resistant security signals that the bank sees distributed-ledger infrastructure as a lasting part of the technology it will need to understand and, where appropriate, build on.
3. How large is HSBC's AI team in Hong Kong?
According to Liao, the bank already employs about 150 AI specialists in Hong Kong. He said that number is expected to double over the next two to three years. Such a headcount places the city among the largest concentrations of AI talent inside the bank, and the growth target suggests it is meant to stay that way.
The size of the team matters because the institute will not start from scratch. It will sit on top of an existing group with experience in building and testing AI systems inside a heavily regulated institution. Adding a research function to a team of that scale is a different proposition from creating a small laboratory with a handful of academics, and it should allow the work to move more quickly from experiment to practical use.
Doubling a specialist team in two to three years is also a recruiting task. Hiring data scientists, machine-learning engineers and cryptographers is competitive across the financial sector and beyond. A named institute with a defined research agenda can serve as a calling card for candidates who want to do more than maintain existing systems, and it may help the bank attract and retain people in a crowded labor market.
Scale brings management questions as well. A team that doubles must keep its standards for model testing, documentation and oversight as new members join, particularly in a bank where each system can affect customers and where supervisors expect clear accountability. The institute's research function may help here by giving the group a shared agenda and a place where methods are examined before they reach everyday banking operations.
4. How does the Singapore centre fit in?
In late July, HSBC announced plans to establish a global AI centre of excellence in Singapore with about 100 specialists. The decision prompted the Hong Kong Monetary Authority to raise questions about the plan, according to coverage of the exchange. The Hong Kong institute, announced afterward, can be read in part as a response, since it commits the bank to a larger and more research-oriented presence in the city.
The figures show why the two projects are not directly comparable. Singapore's centre is planned at around 100 specialists, while Hong Kong already has about 150 and is aiming for 300. The two sites also carry different labels, a centre of excellence in one case and a research institute in the other, which hints at a possible division of focus even though the bank has not described how the work will be split.
For a global bank, operating several technology hubs is normal. Talent, regulation, time zones and client needs differ from city to city, and banks often distribute specialist teams accordingly. What stands out here is that the Hong Kong supervisor publicly queried the choice of another location, and that the bank then followed with a significant commitment at home. How the responsibilities of the two sites develop will be one of the more informative details to watch as both teams grow.
5. Why do post-quantum cryptography and blockchain belong on the list?
Post-quantum cryptography is of interest to every institution that stores sensitive information. Banks hold account records, payment instructions and identity documents, and some of that data needs protection for decades. A concern commonly raised by security researchers is that encrypted data collected today could be stored and decrypted later, once quantum computers mature. Moving to quantum-resistant methods ahead of time reduces that exposure.
The subject also matters to digital-asset markets. Public blockchains depend on digital signatures to prove who owns a wallet and to authorize transactions. If signature schemes were weakened by quantum computing, the security model of those networks would be tested. Research into quantum-resistant approaches therefore has relevance for anyone building on or holding assets on that kind of infrastructure, which is part of why the combination of blockchain and post-quantum work in a single institute is notable.
HSBC has prior experience in this area. In 2024 it tested the interoperability of its tokenized gold product using post-quantum cryptography, an experiment that paired a digital-asset product with quantum-resistant security. The institute will presumably build on that kind of work, although the bank has not said which projects will be carried over or what its first research programs will be.
6. What does this mean for Hong Kong's financial technology sector?
For Hong Kong, the clearest benefit is talent and visibility. A research institute attached to a major bank draws attention from universities, start-ups and engineers, and an expanded team of up to 300 specialists adds a substantial number of skilled jobs. The city has promoted itself as a center for financial technology and digital assets, and a commitment of this kind supports that positioning with a concrete employer and a concrete timeline.
Regulators are also part of the picture. Hong Kong's authorities have shown interest in how banks adopt AI, and a research institute operating in the city gives supervisors a closer view of the work. A research function based locally can make it easier for the bank and the regulator to discuss controls, testing standards and risk management as new tools move from the laboratory into production.
For the digital-asset sector, the signal is modest but real. A large bank has put blockchain on a short list of three research priorities, in a city that is courting the industry. That does not amount to a product announcement, and the bank has given no details about tokens, networks or services. It does indicate that blockchain remains part of the long-term technology planning of a global institution.
7. What should readers watch next?
The institute is expected to begin operations in early 2027, so the first checkpoint is whether it opens on schedule and what team it starts with. The bank has said that the Hong Kong AI team of about 150 should double within two to three years, which gives a rough benchmark against which hiring progress can be measured over time.
The next question is output. Research institutes are judged by what they produce, whether that is published findings, tools used inside the bank, or contributions to industry standards. The bank has not named any specific deliverables for its AI applications, post-quantum or blockchain work, and the first concrete projects will show how the stated priorities translate into action.
Finally, the relationship between Hong Kong and Singapore deserves attention. With a centre of about 100 specialists planned in Singapore and a team heading toward 300 in Hong Kong, the way the bank divides work, reporting lines and investment decisions between the two will show how it balances the interests of two major financial hubs and the supervisors in each.