The global financial landscape is currently undergoing a radical transformation as illicit actors increasingly weaponize artificial intelligence to scale their operations, creating a new sense of urgency within the compliance sector. This shift has moved the industry discourse beyond the theoretical necessity of evolution toward a pragmatic focus on the speed and efficacy of defensive deployment. To address these burgeoning challenges, a cohort of Singapore’s preeminent industry leaders recently convened for "AML Unplugged," an exclusive fireside chat hosted by ComplyAdvantage. Led by Iain Armstrong, Executive Director of FCC Strategy, the event served as a critical platform for the unveiling of the Asia-Pacific findings from the sixth annual State of Financial Crime 2026 report.
The gathering featured a panel of experts including Yi Liang Lee of RazorPay Singapore, Angela Ang from TRM Labs, and Julia Chin of JFourth Solutions. Their discussion traversed the spectrum of modern financial crime compliance, from the persistent, foundational hurdles of data management to the dual-edged sword of artificial intelligence. As financial institutions (FIs) navigate an environment characterized by sophisticated cyber-enabled fraud and complex money laundering schemes, the insights shared at this event provide a roadmap for the future of Anti-Money Laundering (AML) and Countering the Financing of Terrorism (CFT) frameworks.
The Persistent Hurdle of Data Fragmentation
One of the most significant revelations from the panel was that for many financial institutions, the primary obstacle to modernization is not the acquisition of new technology, but the weight of legacy infrastructure. Data fragmentation remains a primary operational hurdle, preventing organizations from achieving a unified, "360-degree" view of risk. Historically, the industry approach was to attempt massive, costly data consolidation projects aimed at creating a single "source of truth." However, the consensus among Singapore’s leaders is shifting toward a more modular and realistic strategy: interoperability.
The panel noted that instead of trying to merge every disparate database into a monolithic whole, firms are now focusing on making their systems "talk" to each other. This shift acknowledges the reality of modern corporate structures, where mergers, acquisitions, and rapid regional expansions often result in a patchwork of software and data silos. The core drivers behind this fragmentation are twofold: the technical debt inherent in older systems and the regulatory complexities of operating across multiple jurisdictions, each with its own data residency and privacy requirements.
From a strategic standpoint, the move toward system coexistence rather than total consolidation allows for greater agility. By using APIs and middleware to bridge gaps between legacy systems and modern AI-driven tools, firms can achieve a comprehensive risk profile without the multi-year downtime associated with full-scale digital overhauls. This approach is particularly vital in the Asia-Pacific region, where the financial ecosystem is highly fragmented and characterized by diverse regulatory expectations.
Separating AI Hype from Operational Reality
The enthusiasm for artificial intelligence in the compliance sector is backed by significant capital and government backing. According to the State of Financial Crime 2026 report, 100% of survey respondents in Singapore indicated that AI-related projects are now more likely to receive budget approval than non-AI initiatives. This trend is bolstered by the Singaporean government’s commitment to invest $1 billion into the AI industry over the next five years. This top-down mandate is filtering through the market, pressuring compliance officers to integrate machine learning and generative AI into their workflows.
However, the panel cautioned against viewing AI as a "silver bullet." The true value of AI lies not in its novelty, but in its capacity to solve problems that are fundamentally beyond human scale. For instance, a human compliance officer cannot realistically synthesize and analyze fifty years of historical transaction data to identify subtle patterns of illicit behavior in real-time. AI transitions from a buzzword to a critical tool when it is used to provide context and insights that would otherwise be lost in the sheer volume of modern financial data.
The panel emphasized that while AI can process 100% of available data, it must function as an "assistant" rather than a replacement for human judgment. This is especially true as global regulators, including the Monetary Authority of Singapore (MAS), place increasing emphasis on "explainability." Financial institutions must be able to demonstrate how an AI reached a specific risk conclusion. Without this transparency, AI-driven decisions could lead to regulatory friction or the inadvertent exclusion of legitimate customers from the financial system—a phenomenon known as "de-risking."
Redefining the Workforce: From Automation to Augmentation
As AI takes over repetitive tasks such as initial screening and basic transaction monitoring, the role of the compliance professional is undergoing a profound evolution. A common fear in the industry is that AI will lead to the mass elimination of entry-level roles. However, the panel argued that such a move would be a strategic error. The traditional "over-the-shoulder" learning model, already strained by the shift to remote and hybrid work, is essential for developing the next generation of compliance experts.
The forward-thinking approach involves transforming junior roles from manual processors into "human-in-the-loop" operators. In this new paradigm, employees are tasked with interacting with, training, and auditing the AI systems. This creates a feedback loop where human curiosity and creativity are used to refine the technology’s accuracy. The panel identified four "Cs"—compassion, curiosity, creativity, and communities—as the essential human traits that AI cannot replicate. These traits are becoming the cornerstone of a more effective, AI-augmented compliance function.
By retaining entry-level staff and pivoting their responsibilities toward AI management, firms ensure they have a pipeline of talent that understands both the technological tools and the ethical nuances of financial crime prevention. This evolution is critical for maintaining a robust defense against "agentic" threats—AI systems used by criminals to autonomously find and exploit vulnerabilities in banking networks.
A Chronology of Compliance Evolution
To understand the current state of financial crime, it is necessary to look at the timeline of technological and regulatory shifts that have led to the 2026 landscape.
- 2015–2019: The Digital Awakening. Financial institutions began moving away from manual, paper-based checks toward basic automated screening tools. However, these systems were often rigid and produced high rates of "false positives."
- 2020–2022: The Pandemic Catalyst. The global shift to digital banking during the COVID-19 pandemic accelerated the adoption of remote onboarding and electronic Know Your Customer (eKYC) protocols. Criminals responded by refining digital fraud and identity theft techniques.
- 2023–2024: The Generative AI Explosion. The public release of advanced Large Language Models (LLMs) gave both compliance officers and criminals access to sophisticated tools for generating content, code, and deepfakes. Regulators began drafting the first comprehensive AI governance frameworks.
- 2025–2026: The Age of Augmentation. The industry reaches a tipping point where AI is no longer optional. The focus shifts to "Agentic AI" and the integration of diverse data sets to combat cross-border financial crime in real-time.
Supporting Data and Economic Impact
The scale of the challenge is reflected in the numbers. Estimates from the United Nations Office on Drugs and Crime (UNODC) suggest that between 2% and 5% of global GDP is laundered annually—amounting to as much as $2 trillion. In Singapore, a global financial hub, the stakes are even higher. The $1 billion investment by the Singaporean government into AI is a direct response to the need for advanced defenses to maintain the integrity of its financial markets.
Furthermore, the State of Financial Crime 2026 report highlights a significant shift in budget allocation. While overall compliance budgets are growing at an average of 10-15% year-on-year, the portion of that budget dedicated specifically to technology and data integration is growing at double that rate. This reflects a realization that human-only teams can no longer keep pace with the volume of digital transactions, which are projected to grow by 20% annually through 2030.
Broader Implications for the Global Ecosystem
The findings from the AML Unplugged event have implications that extend far beyond Singapore. As one of the world’s most strictly regulated and technologically advanced financial centers, Singapore often serves as a bellwether for global trends. The emphasis on "human-in-the-loop" AI and system interoperability is likely to become the standard for financial institutions worldwide.
Moreover, the discussion underscored the importance of public-private partnerships. Initiatives like the MAS COSMIC (Collaborative Sharing of ML/TF Information & Cases) platform demonstrate how shared data can be used to identify criminal networks that move between different banks. For AI to be truly effective, it requires access to high-quality, diverse data sets, which can only be achieved through greater cooperation between competing institutions and their regulators.
Ultimately, the future of financial crime prevention will be defined by how well organizations can balance the efficiency of machines with the ethical oversight of humans. As illicit actors continue to innovate, the compliance industry must not only match their speed but exceed their sophistication by building systems that are resilient, transparent, and integrated. The journey toward 2026 and beyond will be marked by this continuous race, where data is the fuel and AI is the engine, but human intelligence remains the navigator.







