Central banks are cautious of potential risks from artificial intelligence and have yet to widely adopt it, according to a survey by the Official Monetary and Financial Institutions Forum (OMFIF).
Around 61% of respondents from 10 central banks said AI was not supporting their operations. Of those who are using AI, most reported that it was being used for data analysis or data collection.
“Most institutions are experimenting with AI only in low-risk processes: scanning market news, flagging anomalies and summarising reports,” said OMFIF senior economist Yara Aziz.
“Those dipping their toes in describe it as a practical convenience rather than a strategic tool.”
According to OMFIF, the institutions with the most advanced adoption of AI are the most concerned about its risks, citing AI models’ ability to misinterpret unusual data and potential cybersecurity issues.
Nearly 40% of respondents said they planned to upgrade their use of AI.
Around 93% said they did not invest in digital assets like cryptocurrencies, and the vast majority are not seeking to expand into the blockchain realm.
The report also found that almost 60% of respondents plan to diversify their reserves away from the U.S. dollar in the next one to two years, due to faltering trust in the U.S. policy environment. Still, 92% of respondents said the U.S. Treasury market was sufficiently liquid.
U.S. President Donald Trump’s threats to fire Federal Reserve chair Jerome Powell and attempts to remove Fed Governor Lisa Cook have undermined the Fed’s independence this year, with most reserve managers saying the Federal Reserve’s independence was at risk in a July UBS Asset Management survey.
Related content
