Compliance
Ex-MAS Head Warns About Agentic AI Risks With Client Money – Report
.jpg)
The comments by Ravi Menon in Singapore this week add to debate on the proper limits of "Agentic AI" and where its use might affect the fiduciary and other responsibilities of wealth advisors and other financial actors.
The former head of the Monetary
Authority of Singapore, the Asian city-state’s central bank
and principal regulator, has reportedly warned that banks and
wealth managers must set strict limits on the autonomy of AI
agents in managing clients’ money.
Ravi Menon, now chairman of the Global Finance and Technology
Network (GFTN), told a conference this week: “When a family
office appoints an external discretionary manager, it does not
approve every trade. Why would we grant an AI agent looser
boundaries than a seasoned portfolio manager?”
The report was carried in The Business Times (of
Singapore).
AI agents could bring significant gains to investors, but their
ability to make decisions and execute transactions independently
and quickly introduces new risks that existing financial
safeguards may not be able to contain, Menon said.
The comments, issued at an HSBC summit for wealth managers in the
jurisdiction, add to debate about the extent to which AI can
become part of the decision-making process in wealth
management.
Menon stood down as managing director of MAS in 2023.
Separately, Minister of State for Foreign Affairs and National
Development Alvin Tan reportedly told the same HSBC conference
that Singapore must harness AI to generate new wealth and
strengthen its position as a global financial centre, even as it
maintains strong safeguards to protect the integrity of its
financial system. Tan said (source: Straits Times, 17
November) that Singapore does not seek to compete with global
technology giants in developing cutting-edge AI models, but
positions itself as a centre for the research, development and
application of AI across industries, including financial
services.
(Editor’s note: Artificial intelligence has become such a
dominant theme in the wealth sector this year that the issue is
how, rather than if, AI is embedded in the industry’s value
chain. Questions arise as to how the use of AI agents affects the
fiduciary responsibility that wealth advisors have to their
clients, for example. Without a continuing “human in the loop,"
it is unclear whether a firm employing AI will be able to contain
reputational and legal risks without imposing controls. See a
commentary
here.)