Asia-Pacific banks push purpose-built AI adoption for compliance and fraud operations
Asia-Pacific lenders are sharpening their view on how artificial intelligence should be deployed in regulated financial services. A regional survey shows most senior banking executives see specialized AI models as essential for compliance, transparency and faster use in fraud, underwriting and other core operations.
Highlights
- 77% of Asia-Pacific senior banking leaders prioritize purpose-built financial services AI tools for compliance, according to FICO’s August poll of 51 executives.
- 46% expect deployment of task-specific AI in fraud, compliance, and underwriting within one to two years, with 26% reporting systems already in place.
- Trust and governance, especially data privacy and model transparency, outweigh cost as key AI adoption barriers, with 64% citing credit underwriting as the main beneficiary.
Survey findings and deployment outlook
As reported by BusinessWorld, citing FICO, 77% of senior banking leaders in Asia-Pacific say it is extremely important that AI tools used in banking are built specifically for financial services rather than adapted from consumer applications.The poll was conducted in August during FICO’s Banking Leaders Forum in Bali and gathered responses from 51 senior executives and C-suite leaders from major banks across the region. Only 23% say such specialization is somewhat important, underscoring a clear preference for sector-specific systems in a heavily regulated industry.
Dattu Kompella, Asia-Pacific managing director for FICO, says banks operate under a level of regulatory scrutiny that most consumer technology companies do not face. He says lenders need AI models that are explainable, auditable and designed from the outset for compliance.
On deployment timelines, 46% of respondents expect task-specific AI models to be introduced in fraud, compliance and underwriting operations within the next one to two years. Another 26% expect adoption in three to five years, while 26% say their institutions have already deployed such models.
Risk priorities and industry impact
The main obstacle to wider AI adoption is not cost, according to the survey, but trust and governance. Data privacy and security rank as the top concern for 42% of respondents, while 21% cite regulatory and compliance risk and another 21% point to model transparency and explainability; only 16% identify computing cost or infrastructure as the main barrier.For back-office use cases, 64% of respondents say credit underwriting and lending decisions would benefit most from purpose-built AI. Regulatory compliance monitoring is named by 18%, customer due diligence by 10%, and internal policy and business operations review by 8%.
More than 60% of banking leaders also identify fraud and scam detection as the area where generative AI could most improve operations over the next 12 months. That is followed by underwriting decisions at 21%, customer-facing digital banking services at 15%, and regulatory compliance monitoring at 3%, reflecting how banks in the region are prioritizing practical, controlled deployments tied to measurable operational and risk-management gains.
In our earlier coverage, we examined the Bank of England’s call for banks to strengthen cyber resilience by developing “bare metal” recovery options that can keep critical services running if core systems are compromised. The article highlighted regulators’ concern that AI is increasing the scale and complexity of cyber attacks, making end-to-end recovery planning and stronger operational safeguards a growing priority for financial stability.