Financial Crime Compliance Demands Operating Models Before AI

Dr Sebastian Hetzler, Co-CEO of Swiss RegTech firm IMTF, told a Hubbis webinar on 1 July 2026 that AI is the wrong starting point for fixing financial crime compliance. The deeper fix, in his view, is organisational: banks must connect their siloed anti-financial crime (AFC) teams before any new tool, AI included, can deliver. Singapore, he said, is showing what that connected approach looks like in practice.

The argument lands as criminal typologies have shifted from slower, linear patterns to industrialised fraud, deepfake-enabled scams and money-laundering-as-a-service. Hetzler’s case study is IMTF’s Siron®One platform, built around the same convergence thesis and used by more than 1,500 financial institutions globally, per the company and the FinCrimeTech50 2025 report. His wider point: when AFC is run by org chart, the threat picture stays fragmented and AI optimises a broken structure.

The Convergence Argument

Hetzler has spent more than 15 years working in anti-financial crime, first as managing director of Tonbeller, then as FICO’s vice president of product management for compliance solutions. He joined IMTF as Co-CEO in December 2022, when the Swiss firm acquired the Siron business from FICO and pulled compliance, onboarding and customer lifecycle management into one stack.

From IMTF’s perspective, the central argument is convergence. Customer onboarding, KYC, transaction monitoring, transaction screening, sanctions screening and fraud are not separate problems with separate solutions. “We have been proponents of an anti-financial crime convergence, or holistic approach to financial crime,” Hetzler said at the Hubbis Digital Dialogue on holistic AFC, held on 1 July 2026. “Not to look at KYC, customer onboarding, transaction monitoring, transaction screening, sanction screening, fraud, and so on, in silos, but to connect the data.”

That convergence thesis is the design principle behind IMTF’s Siron®One platform and integrated AFC modules. The platform aggregates client, transaction and behavioural data into a single view so that an alert flagged in transaction monitoring can be cross-referenced against onboarding history, adverse media and counterparty exposure in one case file.

IMTF also reports the scale that gives the argument weight. The firm lists more than 1,000 customers served from seven offices across 90-plus countries, with roughly 300 anti-financial crime specialists and a stated 37 years of compliance technology expertise. Siron itself is used by more than 1,500 financial institutions globally, a figure repeated by Hetzler across 2024 and 2025 and corroborated by the FinCrimeTech50 2025 report.

Why Singapore Sits Ahead

Hetzler singled out Singapore as the jurisdiction that has pulled furthest ahead on holistic anti-financial crime. “Singapore has a kind of forerunner position in this market,” he told the Hubbis audience. The reasons, in his telling, sit outside the technology itself: public-private data sharing between financial institutions, regulatory openness to AI adoption, and a supervisor that the industry treats as a partner rather than an adversary.

The shift has substance behind it. Singapore’s MAS has moved to deploy AI against live bank data for scam detection and other financial-crime use cases, deepening its use of machine learning in supervisory work. According to coverage of the programme, MAS is training AI on live bank data to strengthen anti-financial-crime work, and the country has also passed legislation enabling COSMIC, a digital platform that lets financial institutions share information on customers showing potential financial-crime concerns.

Hetzler was careful not to generalise the Singapore pattern across Asia. Different markets, he said, are taking different approaches, and other jurisdictions have taken a stricter line on AI in regulated use.

The distinction matters because AFC transformation is not happening in a vacuum. How quickly a firm can move from siloed compliance processes to wider risk intelligence is shaped by the regulatory architecture around it, and Singapore’s combination of public-private cooperation, data sharing and AI openness is what puts it ahead in Hetzler’s framing.

The Organisational Hurdle, Not the Technology One

This is where Hetzler’s argument becomes harder for the industry to absorb. Most banks, he said, still organise AFC around large specialist teams: one for onboarding, one for continuous KYC, one for transaction monitoring, one for sanctions screening, and so on. Each function tends to have its own tools, its own process, its own culture and its own budget.

The problem is that financial crime does not respect those internal boundaries. A suspicious pattern may sit across onboarding history, adverse information, transaction behaviour, sanctions exposure, counterparty networks and external intelligence. If those signals stay separated across teams and platforms, the institution sees fragments of the risk picture rather than the case.

If we talk about holistic anti-financial crime, I think this is one of the hurdles that we have to overcome. Institutions should not tackle financial crime according to the org chart, but according to the typologies now present in the market.

The quote, from Hetzler during the webinar, captures the operating-model gap. The implication goes beyond tooling: holistic AFC requires a rethink of process, organisation, data and culture together. Buying a new platform or adding an AI model to an existing workflow, on its own, will not deliver the convergence he is describing.

Asked how that level of change gets driven, Hetzler pointed at the top of the institution. Board-level sponsorship is what turns the convergence argument from a slide deck into an operating decision, because moving resources, authority and metrics across AFC functions is the kind of shift that mid-level compliance teams cannot make on their own.

The Threat Picture Has Changed

Hetzler’s wider point is that the pressure for convergence is being driven by what is hitting the banks, not what they have bought. “What has been appropriate in former times is definitely not appropriate to the things we are seeing in the market,” he said, listing the new typologies he sees in client conversations and IMTF’s own deployment data.

The criminal landscape now includes:

• Fast payments and instant settlement, which compress the window in which a transaction can be reviewed.
• Organised cross-border syndicates that span onboarding, payment and exit in one typology.
• Deepfakes and AI-enabled impersonation, opening fraud vectors that bypass traditional identity controls.
• Crypto-related channels, which add what Hetzler called a shadow world for financial transactions that the institution’s AFC framework has to include in its risk picture.
• Fraud-as-a-service and money-laundering-as-a-service, industrialising what used to be bespoke criminal work.

“The Hubbis event listing for the 1 July 2026 webinar placed Hetzler on a panel alongside Kelvin Chiang of Bank of Singapore and Alwyn Loh of KPMG Singapore, both of whom have written about the same shift in client-side practice. The implication Hetzler drew is that AFC frameworks built for slower, more linear and more easily categorised risks are not sufficient. Financial institutions are now in an arms race against more professional, more digital and more automated criminal groups, and the gap between institutions, systems and teams is what those groups exploit.

AI as an Enabler, Not a Substitute

Within that picture, Hetzler was clear about where AI sits. “AI is an enabler,” he said, framing its value as the ability to bring intelligence into the process, process large volumes of data and connect that data across use cases. He expects the next twelve months to bring broader AI adoption in compliance and a continued shift away from siloed organisations.

But he drew a hard line on what AI cannot do. If institutions keep the same siloed structure and use AI to optimise each legacy workflow in isolation, the technology will deliver incremental gains, a modest reduction in false positives here, faster alert triage there, but it will not address the real shift in criminal typologies. The larger risk is that firms use AI to optimise the old model and miss the change that has happened underneath it.

From IT Project to Board-Level Mandate

Hetzler closed the conversation with a transformation message that reached beyond the technology pitch. IMTF can help from a technical standpoint, pulling different AFC use cases and signals into one platform and using AI to connect data and add an intelligence layer.

But he was explicit that technology alone is not enough. “At the end, it is more than just IT,” he said. “IT is important, technology is important, but changes have to start, maybe at the board level.” For an industry that has historically framed compliance investment as an IT or risk-budget decision, that is a different conversation.

The firms that make the shift will be better placed to detect patterns across AFC disciplines, adapt to new typologies and use AI as a genuine intelligence layer. Those that stay in the old model may gain efficiency, but not resilience. For Hetzler, the direction of travel is already set: the future of financial crime compliance is connected intelligence, not isolated controls.

Frequently Asked Questions

What did Dr Sebastian Hetzler argue at the 1 July 2026 Hubbis webinar?

Hetzler, Co-CEO of IMTF, argued that AI alone cannot fix financial crime compliance while banks still organise their AFC teams in silos. The required shift is organisational first: connecting onboarding, KYC, transaction monitoring, sanctions screening and fraud data across teams before adding AI on top.

Why is Singapore seen as a forerunner in anti-financial crime compliance?

Singapore has built the regulatory and data-sharing architecture that supports convergence. MAS has moved to deploy AI against live bank data for scam detection, as covered in reporting on the Singapore MAS AI programme for live bank data, and legislation has enabled Singapore’s COSMIC platform for cross-bank crime data. The combination of public-private cooperation, data sharing and AI openness is what puts Singapore ahead in Hetzler’s framing.

What does holistic anti-financial crime compliance mean in practice?

Holistic AFC means treating KYC, onboarding, transaction monitoring, transaction screening, sanctions screening and fraud as one connected risk picture rather than separate functions with separate teams, tools and budgets. Signals that span those functions, such as a customer whose onboarding history, transaction behaviour and counterparty exposure together suggest risk, become visible only when the data is integrated.

Will AI fix financial crime compliance on its own?

Hetzler’s answer is no. AI can process more data, connect intelligence across functions and help with alert prioritisation, but applied to a still-siloed operating model it delivers incremental efficiency rather than a structural improvement. The board-level organisational change, breaking the org chart between AFC teams, has to come first.