United Overseas Bank and the consultancy Accenture have signed a memorandum of understanding that lets the bank adopt generative and agentic artificial intelligence across three years. The two describe it as the first such collaboration between Accenture and a Singapore bank, and the bank says it will lean on the consultancy's AI platform and its training service for staff.

The framing is deliberately broad. Wee Ee Cheong, UOB's deputy chairman and chief executive, said the technologies hold out the prospect of personalised services delivered quickly across the bank's ASEAN network, and Julie Sweet, Accenture's chair and chief executive, said the aim is to scale generative and agentic AI quickly. UOB, incorporated in 1935, operates more than 470 branches and offices in nineteen markets.

What a memorandum commits anyone to

A memorandum of understanding is a statement of intent, not a contract for a delivered system, and this one comes without a figure attached. There is no value, no timetable and no named project. The models the bank intends to deploy are also among the ones whose failure modes are least understood: an agent that can act on a customer's account is a different proposition from a chatbot that suggests a savings plan, and the risk sits in the same software the bank is advertising.

That does not make the deal hollow. A training arm is a real asset for a workforce that has to use these tools, and a vendor platform is a way to start without building from nothing. But the measure a reader should hold is annual: how many processes the bank actually runs on agents, and what it removed as a result.

The release offers neither, and a three-year horizon is long enough to hide the answer for a while.