Using AI to Streamline Compliance Processes

2024 is proving another standout year for the regulatory space, finding itself under the spotlight, for better and worse reasons. This month, The Fintech Times will look at some of the biggest issues regarding compliance and financial rules, as well as the solutions hoping to ease the compliance journey for firms and make the fintech world fairer and safer. FullCircl's Banking Success Director, Lucy Huntley, talks to the FinTech Times about the AI revolution in compliance and how it can be harnessed for growth.

Having already explored compliance challenges, penalties and solutions, we now turn our attention to the technology of the moment: AI.

While we’re well aware that AI is currently spoken about in absolutely every context, we also understand the huge impact it can have across sectors and operations. With this in mind, The Fintech Times reached out to industry experts to ask how AI will leave its mark on compliance for the fintech industry. Read the original FinTech Times article here.

AI promises ‘massive impact’ on compliance efficiency

Now is hardly the very beginning of AI supporting compliance processes, explains Lucy Huntley, banking success director at FullCircl. However, its potential impact on the space is yet to be fulfilled.

“In many respects, AI has been playing a significant role in automating compliance processes for a long-time now. AI has already transformed compliance processes in the fintech industry by making them faster, more accurate and more efficient. With AI, tasks including regulatory reporting and disclosure, data analysis, and risk assessments can be automated, saving time whilst also reducing errors and improving the customer experience, particularly at the onboarding stage.

“Likewise, machine learning algorithms continuously learn from data, improving accuracy over time. This means compliance teams can focus on more strategic tasks, while AI will always look to keep the company compliant!

“Looking to the future AI will have a massive impact on improving automation processes around document verification. Computer Vision (CV) for example is an emerging field of AI which enables a much deeper level of accuracy when cross-referencing the biometrics of a live selfie with a portrait on an ID document.”

“AI adoption in compliance is still in its early days, but is rapidly picking up speed,” explains Paul Cottee, director of compliance SME at NICE Actimize. “Currently, the adoption of AI centres around assisting with often mundane tasks: that is, for example, using a large language model to help speed up written tasks such as reports and filings; using AI to help score and prioritise surveillance alerts; and looking for patterns and relationships across large volumes of unstructured data.

“This is increasing the efficiency of the compliance function, both in terms of monitoring and supervisory activities, and also in administrative work, leaving the compliance officers with more time to devote to tasks requiring human judgment and decision-making.”

Saving compliance teams time

Hilary Wandall, chief ethics and compliance officer at business intelligence and data firm Dun & Bradstreet, also echoes this sentiment. She explains while AI can streamline compliance processes and save compliance teams time to spend elsewhere, firms need to be careful about how they implement it.

“AI is beginning to play a pivotal role in streamlining certain regulatory compliance processes within the fintech industry. The emergence of AI and ML tools has enabled companies to analyse vast amounts of data in real time, detecting patterns that indicate potential compliance risks, such as money laundering, sanctions, or fraud.

“This improves customer due diligence by concurrently cross-referencing many databases to verify identities with potential clients and any risks associated with them. The impact is profound, as AI-driven automation reduces costs, minimises human error, and enhances the speed and accuracy of compliance tasks. As compliance teams face increasing workloads due to new regulations, automation tools allow them to focus on more complex or ambiguous cases.

“However, the effectiveness of AI models primarily hinges on a business’s robust understanding of its data estate and the implementation of an adequate data governance system. AI is only as smart as the data that fuels it, making it imperative to introduce policies and adhere to data quality standards. Incorporating high-quality, comprehensive data enables teams to assess risk more accurately and make informed decisions, leading to proactive risk management and fostering innovation and competitive advantage in the industry.”

Assessing the risk

For Steve Bradford, senior vice president of EMEA at SailPoint, firms planning to implement AI into their compliance processes must be willing to ensure they have the correct amount of oversight on its use, as well as safeguard data to a high level.

“Fintechs should put risk analysis processes in place to help ensure regulatory compliance and prevent situations that could lead to fraud or data leakage. Through a unified, AI-enabled approach to identity security, organisations can ensure that staff have only as much access as is required to perform their assigned roles and responsibilities – no more, no less.

“Using AI speeds and streamlines identity decisions, something crucial given the pace at which businesses – and cyber threats – are evolving. This enables security teams to move faster and more effectively to spot and stop unnecessary, inappropriate, or potentially compromised access.

“Safeguarding data is business critical. With the stakes higher than ever before, fintechs must make full use of the available AI-driven tools and technology to gain better visibility and insight into the specific risks associated with user access. A carefully considered approach to identity security, with stringent policies on how access to data is managed and controlled, will help businesses keep compliant as well as stay one step ahead of cybercrime.”

Saving time and money

Bronwyn Boyle, CISO at PPRO, explains how the use of AI could save not only resource, but also reduce costs – although warns that this isn’t guaranteed: “AI is bringing multiple benefits to the fintech industry by automating compliance processes.

“Real-time monitoring of transactions and activities dramatically improves the time to detect potential compliance, security, or fraud issues, while machine learning enhances the ‘signal to noise’ ratio by correlating data points across multiple sources and reducing false positives.

“AI can help businesses save time, allowing fintechs to pool resources toward innovation and growth initiatives instead.

“While early cost reduction opportunities seem promising, the longer-term impacts on cloud computing costs and ESG footprints are yet to be fully understood. However, at this stage, businesses are expected to hire in new areas like AI management and compliance to ensure the technology is being used appropriately and in accordance with regulations.”

Streamlining processes

“As more organisations implement AI for compliance, one thing we are seeing is a reappraisal of the risk management process and the role of the compliance team,” explains Joel Lange, EVP and general manager for Dow Jones Risk & Research.

“Research is time-intensive and expensive, but with AI, processes such as negative news screening can be conducted more quickly than ever before. This means that compliance can be accelerated within the decision-making workflow, enabling organisations to assess at the very beginning of a potential relationship whether it should be pursued or not.

“This avoids the wider organisation wasting time and effort into developing a business opportunity that should never have got off the ground and helps to eradicate the stereotypical view of compliance as a blocker on activity.”

AI’s growing role in compliance

Finally, Gabriel Hopkins, CPO at Ripjar, weighs up just how influential AI could be in changing the face of compliance for fintechs: “AI is not a panacea, but it does provide some powerful tools to make lives easier.

“While much of the AI we talk about today is generative AI, established AI and ML capabilities also play an important role and teams should be careful to find the right tools for the right tasks. Historically, teams have been slow to adopt AI because of fears about predictability and explainability in a regulated setting. However, well-implemented models can have a significant difference in compliance activities, particularly when dealing with complex matching challenges.

“The potential for generative AI is even greater. Some uses are obvious. The drafting of narrative summaries for Suspicious Activity Reports (SARs) can be supported by GenAI providing seed summaries which an analyst can quickly validate and then submit.

“Maybe more exciting is the chance to reinvent how compliance tasks are performed, powering virtual analysts to achieve, and sometimes surpass, human-level accuracy. Indeed, these tools can revolutionise compliance processes and help users save up to 200 hours annually.

“Analysis has shown that a virtual analyst can help with both false positives, which are incorrect risk flags, and false negatives, which there are existing risks that haven’t been flagged – both are endemic issues caused by dealing with very high volumes of matches. GenAI can reduce these inaccuracies, significantly enhancing the efficiency and effectiveness of customer and counter-party screening.

“Compliance teams should continue to ensure that they carefully select the right AI capabilities and carefully validate the results on an initial and ongoing basis to ensure success.”