IFCTF 2026: AI Must Augment Human Judgment, Not Replace It In Financial Crime Fight - Experts

KUALA LUMPUR, Oct 6 (Bernama) -- Artificial intelligence (AI) should be used to strengthen human judgment rather than replace it in the fight against increasingly sophisticated financial crime, with financial institutions also needing to bridge critical AI skills gaps, experts said.

Asian Institute of Chartered Bankers (AICB) chief executive officer (CEO) Edward Ling said that according to AICB’s AI in Practice report, there is growing adoption of AI in areas including fraud detection, know-your-customer (KYC), anti-money laundering (AML) and counter-terrorism financing (CFT), but human confidence in AI-generated outputs for important decisions remained limited.

He said only 25 per cent of respondents in AICB's study said they trusted AI-generated outputs sufficiently to act on them, particularly for key business decisions.

“AI is extremely good at processing volume data, identifying patterns and finding anomalies that humans may miss. But identifying a pattern and understanding what that pattern means are two different things.

“AI should amplify human judgment, not replace it. The machine can tell us where to look. Ultimately, a properly trained professional should be able to challenge the output, understand the context and take responsibility for the decision,” he said.

Edward said this during a media roundtable organised by AICB in conjunction with the 16th International Conference on Financial Crime and Counter Terrorism Financing (IFCTF) 2026 here, held under the theme “The New Face of Financial Crime: AI, Intelligence and the Race for Trust.”

He noted the industry’s ability to harness AI effectively was also being constrained by a shortage of specialised skills, with 79 per cent of respondents in AICB's study reporting gaps in specialised AI technical skills.

According to Edward, the solution was not to turn every banking professional into an AI or data science specialist, but to equip them with the ability to understand, question and responsibly use the technology.

“For most banking professionals, we need sufficient AI and data literacy to understand what technology can and cannot do, the professional judgment to challenge AI output, and a good understanding of governance issues such as ethics, privacy, explainability, bias and accountability,” he said.

Meanwhile, Asia-Pacific (APAC) compliance and financial crime Industry advisor at SymphonyAI Singapore Martin Smith said financial institutions must shift from a reactive, after-the-event approach to a proactive intelligence-driven model to anticipate financial crime threats.

He said transaction activity, customer onboarding, negative news and other information held by banks should be analysed collectively to identify patterns that could indicate financial crime before losses occurred.

Martin said this required financial crime organisations to change not only their technology but also their people and processes, while developing multidisciplinary teams capable of assessing risks faster.

He said criminals were not constrained by the bureaucratic and jurisdictional barriers faced by financial institutions, highlighting the need for ASEAN and international standard-setting organisations to develop mechanisms for more effective cross-border sharing of risk-related information.

At the same time, Prevent Crime Now founder and crime prevention specialist Dr Shamir Rajadurai said institutions should understand how criminals think instead of focusing solely on the latest scam tactics.

He said compliance teams needed to put themselves in the mindset of those seeking to bypass controls, allowing institutions to identify vulnerabilities and intervene before a crime occurred.

Shamir said individuals should not be expected to combat increasingly sophisticated scams alone, as criminals were becoming more adept at exploiting human psychology and using technology to enhance their approaches.

Meantime, Transparency International Malaysia president and GraymatterForensic Advisory Sdn Bhd managing principal Dr Raymon Ram said AI should primarily be used to identify red flags and connect information, rather than make decisions on behalf of institutions.

He said every decision involving AI should remain attributable to a person, while the process and model used to reach the decision must be sufficiently transparent and auditable to ensure accountability.

Confide Platform founder and CEO Pav Gill said the principle of accountability also applied to whistleblowing systems, which should be regarded as an early-warning mechanism rather than merely a response after financial crime had occurred.

Pav said employees would only trust whistleblowing channels if they saw evidence that disclosures were acted upon and that those who raised genuine concerns were protected from retaliation.

He said AI was also creating a new dimension in the fight against financial crime as criminals and institutions increasingly used the technology against each other, requiring organisations to consider how automated systems themselves could be governed and monitored.

The experts agreed that the fight against financial crime could not be left to technology, regulators, law enforcement or the public alone, but required stronger human capabilities, responsible AI, institutional accountability and effective information-sharing.

-- BERNAMA