Artificial intelligence can’t yet be trusted to make decisions on interest rates, a leading bank authority chief has said, claiming the software still lacks judgement skills.
But AI could make it easier for criminals to launch cyber attacks, warned Chia Der Jiun, the head of the Monetary Authority of Singapore.
He said AI was being used in some economic models and in areas such as fraud detection, but stressed it was not at a stage where it could “supplant human judgement”.
“There is a great deal of judgement involved in understanding and having a view as to the forward path of inflation… [and] the transmission of monetary policy,” Chia said during a Bank for International Settlements’ panel session on AI use.
Also on AF: Hang Seng Lifted by Property, Data Boosts; Tech Drags on Nikkei
“Maybe it will happen one day when you can blame the AI for mistakes in monetary policy,” the managing director of Singapore’s central bank joked. “But if a human has to take the blame then the human will have to make the judgement”.
Chia, who took over as MAS chief at the start of the year, also warned the main risk was of AI being used by cyber criminals.
“Another dimension that is not talked about that much is that AI has the potential to democratise – not in a good way – but to democratise access to malware,” Chia said.
There were already tools available online he said, “for people that are not too sophisticated in their technical skills” to use to launch cyber attacks. “It is a space we have to watch.”
- Reuters with additional editing by Sean O’Meara
Read more:
US May Deny China, Russia Access to Advanced AI Models
Apple Working With TSMC to Make its Own AI Chips: WSJ
Warren Buffett Likens AI to the Atomic Bomb — Quartz
US Asks China to Declare AI Will Never Control Its Nuclear Arms
Apple to Focus on China, Roll Out AI iPhones to Revive Sales
