TLDR: AI agents are poised to revolutionize financial services, offering significant productivity gains and enhanced capabilities in areas like fraud detection and customer service. However, their widespread adoption and effectiveness are critically dependent on robust data foundations. Challenges such as data fragmentation, metadata gaps, and cross-border data sovereignty rules pose substantial hurdles, requiring financial institutions to prioritize seamless data integration, transparency, and clear regulatory guardrails to fully realize the promise of AI.
Artificial intelligence (AI) agents are rapidly transforming the financial services landscape, promising unprecedented efficiencies and advanced analytical capabilities. Institutions like the Bank of Thailand are already leveraging AI agents, such as an SQL Coding Copilot, to convert natural language requests into SQL queries, granting non-technical staff access to complex loan-level data. This, combined with metadata copilots and multi-agent reasoning, empowers employees to inquire about lending patterns for unusual activity or compliance gaps and receive accurate, timely answers. The integration of these AI agents has notably strengthened the central bank’s ability to detect money-laundering red flags earlier and bolster overall financial stability.
The appeal of AI agents stems from their capacity to autonomously collaborate on end-to-end tasks, with human intervention reserved for exceptions, oversight, and coaching. An International Data Corp (IDC) survey indicates that 86% of Southeast Asian companies plan to deploy AI agents within the next year. McKinsey forecasts a future where each human practitioner could supervise at least 15 AI agents, potentially unlocking a 20-fold productivity gain. Early adopters, such as Oversea-Chinese Banking Corporation (OCBC), have already seen tangible benefits, reducing private banking onboarding from 10 days and 40 documents to under an hour while simultaneously improving standardization.
Despite this immense promise, scaling AI agents is fraught with challenges, primarily centered around data. Data fragmentation is a significant obstacle, as Remus Lim, Senior Vice President for Asia Pacific and Japan at Cloudera, explains: ‘Transaction records, customer files [and other information] often sit in separate systems spread across on-premises, cloud and regional platforms. Without seamless integration, AI agents struggle to generate timely and accurate insights.’ This issue is compounded by metadata gaps, which are crucial for tagging transactions with essential details like timestamps, currency, and compliance markers.
Multinational banks face an even more acute challenge, as noted by Gavin Day, Chief Operating Officer of SAS. These large institutions often maintain disparate data strategies across different countries and divisions, resulting in a fragmented technological landscape that hinders interoperability. Furthermore, data sovereignty rules frequently prevent the consolidation of data across borders, severely limiting the effectiveness of AI agents.
Beyond data integration, transparency and clear guardrails are paramount, especially as AI agents assume more autonomous decision-making roles. Regulators are increasingly demanding transparency in how AI agents arrive at decisions, strict adherence to data sovereignty, and clear audit trails – requirements that many legacy systems were not designed to meet. Lim emphasizes that ‘Having these [guardrails] in place not only reassures regulators but also builds consumer confidence that AI is being applied responsibly in sensitive areas of banking.’ Escalation thresholds, automated alerts routed to human investigators, and dashboards tracking agent behavior are essential for ensuring accountability and responsible AI deployment.
In Asia Pacific (Apac), the financial impact of AI agents is widely recognized. A Salesforce survey reveals that 75% of chief financial officers (CFOs) believe AI agents could boost their company revenue by nearly 20%, with 77% anticipating a transformation in their business models. Robin Washington, President and Chief Operating and Financial Officer at Salesforce, highlights this shift: ‘With AI agents, we’re not merely transforming business models; we’re fundamentally reshaping the entire scope of the CFO function.’ This has led to 83% of CFOs using AI in decision-making for risk assessments, financial forecasting, and profitability analysis.
Also Read:
- Agentic AI Poised to Revolutionize Finance: Opportunities and Risks Explored by Kyndryl Institute
- AI Agents Poised to Revolutionize Professional Services, Boosting Pipeline Capture Beyond 20%
Ultimately, while AI agents are poised to revolutionize banking, their full potential hinges on a robust data foundation. Without strong data integration, effective sovereignty safeguards, and transparent operational frameworks, the promise of AI agents in financial services risks falling short. The clear message for financial institutions is to prioritize getting their data foundation right first.
