ACAMS Hollywood - Overcoming inertia: the creative ways AML/CFT and Fraud teams adopt the latest regtech
The adoption of advanced financial crime detection technology is paramount for Anti-Money Laundering/Combating the Financing of Terrorism (AML/CFT) and Fraud teams. However, the journey towards embracing the latest regtech tools is not a one-size-fits-all approach. Rather, it requires a shift in mindset and a willingness to explore creative and iterative methods. At this year’s ACAMS Hollywood conference, Hawk AI hosted a panel discussion delving into the innovative ways that financial institutions can overcome inertia and seamlessly integrate regtech into their operations.
The Panelists
- Moderator: Austin Hong, Partner, Oliver Wyman
- Tobias Schweiger, Co-founder & CEO, Hawk AI
- Xan Kasprzak, VP Product, CSI
The state of financial crime detection technology adoption
Austin Hong set the stage by polling the audience. In response to the first question, “What is the largest roadblock to adopting new technology at your organization?” the audience overwhelmingly voted for “lack of IT resources and prioritization.” This suggests that the inertia to technology adoption stems from leadership not seeing it as an organizational priority and allocating funds, staff, and resources accordingly. The responses to the question, “How confident are you in the ability of your organization to effectively implement new technology?” were more mixed, but the audience trended toward the confident end of the spectrum. This again highlights the likelihood that sluggish technology adoption is not a question of ability or capacity, but one of resourcing.
To set the stage, Hong emphasized the remarkable evolution of compliance technology recently, particularly in the last three to five years. The technology has made significant strides, and it is more mature now than ever. At the same time, financial crimes are increasing, regulatory pressure on financial institutions to detect that crime is mounting, and fines for non-compliance are increasing year-over-year. The need for this technology is great, and yet organizations still face inertia to adopting it.
Appetite for cloud and AI/Machine Learning technology
Tobias Schweiger mentioned an increasing demand among financial institutions for cloud-based solutions and AI/Machine Learning technology. The advancements in cloud technology allow for faster implementations and scalability, helping organizations to streamline processes and accelerate their time to market. AI technology can help AML/CFT, and Fraud teams reduce false positives, minimizing the burden of investigating unnecessary alerts and allowing investigators to focus on genuine risks. AI-powered anomaly detection also supports AML/CFT, and Fraud teams by detecting “unknown unknowns.” With advanced analytics, investigators can identify patterns and deviations from normal behavior, enabling the swift reporting of suspicious activity. Many financial institutions are eager to see what these technologies can do. However, despite the clear advantages, the prevailing inertia hinders adoption.
“Applying modern technologies like these is the only way to fight financial crime for good,” said Schweiger.
Obstacles to Implementing new technology for AML/CFT and Fraud
We’ve established that there is inertia to implementing new AML/CFT and Fraud prevention technology. What are the specific obstacles that financial institutions must overcome? One of the greatest challenges lies in selling the idea of implementing new technology to internal stakeholders. Convincing decisionmakers and teams within the organization about the value and benefits of adopting this new tech can be a daunting task. However, the key to overcoming this challenge lies in building a solid business case, centered on the return on investment. When they see how the new technology can improve efficiency, enhance detection capabilities, and reduce financial risks, stakeholders are more likely to support adoption. Additionally, creating a coalition of stakeholders across various departments and functions can help build momentum and address concerns effectively.
A crucial differentiator in AI technology is explainability. Because of the regulatory burden financial institutions carry, the ability to explain the decisions and outputs generated by AI models is crucial. Clear and transparent explanations can help build trust and overcome resistance to the adoption of AI-powered solutions. By addressing these obstacles head-on, organizations can pave the way for successful implementation and unlock the potential of new technology in the fight against financial crime.
“Everyone's going to roll out AI,” said Xan Kasprzak. “Everybody's going to adopt it.” The question then becomes, what AI a financial institution adopts, and how they implement it.
AML/CFT and Fraud system augmentation vs. “rip and replace”
Some financial institutions hesitate to implement new technology because of the cost of switching from legacy systems. In these instances, a system augmentation may make more sense than a full “rip and replace” of the legacy system. There are benefits and drawbacks to both approaches. Augmentation requires a deep understanding of the legacy system, as it integrates new technology with the existing infrastructure. This allows financial institutions to continue to leverage their investment in the legacy system while enhancing its capabilities. Augmentation is feasible only if data can be obtained from the system with minimal disruptions. This can be a relatively easy and effective way to experiment with new technology, allowing financial institutions to test it for potential benefits before committing to a complete system replacement.
Augmentation represents an iterative approach to AML/CFT and Fraud technology adoption, featuring continuous system and process improvement. Financial institutions can try new technologies, analyze results, fine-tune the system, and repeat the process, enhancing performance over time.
“You show again and again that you know the result is correct. That the result is what we all want,” said Schweiger. “That’s why we are very quick and successful when it comes to augmenting a legacy system, which is one part of our solution. Of course, we replace legacy just as much.”
It is important to note that augmentation may have limitations, especially if the legacy system is outdated and lacks compatibility with new technology. A careful evaluation of the legacy system's capabilities and potential for augmentation is necessary before deciding between augmentation and replacement strategies.
How to make the business case for new AML/CFT and Fraud technology
To convince internal stakeholders to implement new AML/CFT and Fraud tech, it is crucial to highlight the business benefits. One significant advantage is increased efficiency. By implementing new technology, financial institutions can refine their legacy systems and streamline their operations, reducing reliance on large teams and manual processes. As Kasprzak aptly put it, the ability to work smarter, not harder, is a fantastic outcome for any business. This is particularly important when considering the staggering volume of alerts that AML/CFT and Fraud teams face. For instance, FinCen and the OCC recently fined USAA for having an insufficient AML transaction monitoring program. The regulators based this action in part on USAA having a backlog of 90,000 alerts. “How do you apply enough people to resolve that?” asked Kasprzak. AI-powered solutions help financial institutions manage and resolve alerts in a timely and accurate manner. It is becoming more necessary to use this technology to remain complaint with regulatory requirements.
Harnessing the power of new technology can significantly reduce friction in the customer experience, particularly via account opening fraud and account takeover fraud mitigation. AI algorithms can swiftly analyze vast amounts of data, including customer information and transaction patterns, to identify potential risks and detect fraudulent activities in real-time. This proactive approach ensures a smoother account opening experience for legitimate customers and protects them from fraudulent use of their account. By implementing AI-driven solutions, financial institutions can strike a balance between robust security measures and seamless customer interactions, ultimately fostering trust and satisfaction among their client base.
Utilizing AI technology for real-time fraud detection and prevention plays a crucial role in safeguarding a financial institution's bottom line. Fraudulent activities can cause significant financial losses, damage reputation, and erode customer trust. By leveraging AI algorithms, financial institutions can quickly analyze vast amounts of data to detect and stop fraudulent transactions in real-time. AI-powered systems can also adapt and learn from new fraud patterns, continuously improving their accuracy and effectiveness in blocking fraudulent attempts. By efficiently identifying and blocking fraudulent transactions, financial institutions can minimize financial losses, avoid costly chargebacks, and reduce operational expenses associated with fraud investigation and remediation.
Running a Successful Pilot
Referencing the initial polling questions, the panelists noted that financial institutions need to ask themselves if they have the resources and ability to implement new technology successfully. Adequate funding, skilled personnel, and infrastructure support are crucial for a smooth implementation. “You need to identify the right stakeholders internally that can speak to the IT and the data science,” said Kasprzak.
Data gathering and conversion play a pivotal role in a pilot's success. Financial institutions need to ensure access to quality data relevant to the AI model's training and testing. This may involve data gathering from various sources, cleaning and preparing the data, and converting it into a format suitable for the models. Accurate and comprehensive data lay the foundation for the pilot's precision and reliability. By addressing these points diligently, financial institutions can set themselves up for a successful AI technology pilot, laying the groundwork for future implementations and potential advancements in their AML/CFT and Fraud prevention efforts.
When it comes to internal data science resources, there’s no one-size-fits-all approach. Larger banks will often have entire data science departments. However, a large data science staff is not always necessary thanks to the no-code configuration available in new AI systems. The necessary data science resources vary, depending on the financial institution and its individual needs.
The role of regulators in AML/CFT and Fraud AI adoption
Regulators play a crucial role in the adoption of AI technology for AML/CFT and Fraud prevention. Regulators can exert pressure on financial institutions to adopt AI more quickly via regulatory requirements, guidelines, and expectations, creating a sense of urgency. Regulators can also provide a supportive environment for Financial Institutions to explore and pilot new technologies. By offering sandboxes and pilot programs, regulators can encourage Financial Institutions to test and validate AI solutions in a controlled setting.
“The room is given to financial institutions from a supervisory standpoint, to try out new technology, to sandbox stuff, to pilot things in a safe environment,” said Schweiger.
Cross-institutional AML initiatives facilitated by regulators promote collaboration and knowledge sharing among Financial Institutions, further fostering innovation and advancements in AI adoption. Regulators also play a vital role in providing guidance on AI adoption, helping Financial Institutions understand the regulatory expectations, risks, and best practices associated with integrating AI into their AML/CFT and Fraud prevention strategies.
In a final word to regulators on the topic, Schweiger said, “I would encourage regulators around the world to open the doors more for technology that will help overcome this inertia.”
“They are,” Kasprzak responded. “They're opening that door and being more receptive and supportive.”
The evolving and supportive stance of regulators toward AI adoption is encouraging, as they recognize the potential of the technology to enhance financial crime prevention efforts. The continuous evolution of regulatory guidance will support the industry in its need for clarity and confidence around AI technology for AML/CFT and Fraud prevention.
The Bottom Line
Implementation of new technology doesn’t have to be all or nothing. There are as many ways to implement AI and machine learning for AML/CFT and Fraud as there are financial institutions. What’s most important is that financial institutions understand the obstacles and overcome the inertia, improving their efforts to detect and stop financial crime.