Blog

Top Trends in Anti-Financial Crime for 2024

02.05.2024 | Charmian Simmons
 

Several key trends are shaping the landscape in 2024 and demanding attention from professionals in anti-financial crime and fraud management. These trends span AI-powered solutions, technology, and resource capabilities, as well as the use of data, regulatory expectations, and a heightened focus on proactive measures.

Are you in the know? In this article, we look at the top trends in anti-financial crime for 2024, focusing on emerging trends to be aware of, as well as trends that retain their place in the list from last year. Prefer video? Watch the accompanying webinar.

The New Trends in Anti-Financial Crime for 2024
More investment and integration with AI

Top trends in anti-financial crime in 2024 don’t get more obvious than this – more investment in and integration of AI. If last year showed the potential of AI technology, 2024 is set to be the year that enterprises see even more value in using it to fight financial crime. Benefits include more comprehensive and faster investigations, up-to-date AI-created case summaries, instantaneous suspicious activity reports pulling together all the information that has been collated on a client, and much more. Backend capabilities include analysis of investigators, automatically assigning tasks, and advanced feedback analytics. With generative AI expected to reach all areas of anti-financial crime (KYC, CDD, AML, fraud, watchlist management) this year, enterprises will need to adopt, invest, and integrate with AI to maximize their potential and save on costs.

2023 saw financial services increasingly adopting AI innovation, such as the Sensa Copilot and Sensa Investigation Hub from SymphonyAI. The trajectory of these use case instances will continue to broaden rapidly, inspiring new developments and ways of using AI to achieve greater results. That said, innovation at pace must be managed and governed. It’s therefore sensible when adopting new AI innovations and platforms – particularly third-party integrations – to perform sound due diligence and be satisfied it works within regulation(s) and creates a safe and secure environment for data processing and storing. The trajectory of these use case instances will continue to broaden rapidly, inspiring new developments and ways of using AI to achieve greater results.

On a broader note, enterprise companies like Adobe, Microsoft, and Google are already integrating AI into product offerings for use across multiple business functions. Expect this to continue in the next twelve months and beyond as many tech firms innovate at pace. In fact, Aperio’s research suggested the AI-related software industry is expected to grow to $26.67 billion by 2026 while other suggest it will reach $76 billion. As such, it is highly likely that use of AI will hit every industry this year to help boost productivity, effectiveness, and efficiency.

Verticalized language learning models (LLMs) to increase

An important evolution in generative AI adoption in 2024 is verticalized large language models (LLMs)—as opposed to horizontal LLMs like ChatGPT—that improve the way we use and consume AI inputs and outputs, especially at the enterprise level for specific use cases. AML and fraud management are prime vertical examples of this advancement and already benefiting from it.

Unlike the explosion in 2023 of generic open-source GPT models trained on every open-source topic available (from dog breeds to complex IT coding practices), at the enterprise level, use-case demand is pushing for industry-specific models that are pre-trained in a specific subject matter and/or industry sector. Comfort is gained in the fact that once in place, they do not require explanation of subject matter context or acronyms commonly used in that setting for the models to show immediate value and results from the get-go and can build end user confidence in the technology.

This is a game changer at multiple levels in financial crime prevention. It allows investigators to research far more quickly than previously, pulling together contextual information by asking the chatbot rather than looking at it themselves. In this way, it will also improve results and save money for the company that uses it; far less time will be spent on each investigation. Alongside this, by using vertical-specific gen AI LLMs, uniformity can be created around case summaries and suspicious activity reports, improving their quality. It isn’t just the front end that benefits either; using such software, enterprises will enjoy improvements in data and feedback analytics, which are then fed back into the LLM, improving it over time.

It’s therefore easy to see why this is a big trend – most visibly, it speeds up productivity by ensuring more accurate results but also offers many other benefits as well. As such, off-the-shelf AI applications for specific verticals (such as financial services, anti-financial crime, fraud, or insurance) will become more prevalent, improving the time to value of software. As such, companies will start seeing the fruits of their investment far more quickly than previously.

In conjunction with the rise of vertical AI software, it’s expected that cloud computing will play a big part in this emerging area; developers will be able to enhance their product and push it to users in line with any regulation changes or requests from clients. This makes it faster to debug the software, adds value, and frees up internal IT teams for other projects.

Gen AI Copilots will boost efficiency and productivity across the board

The integration of gen AI copilot technology powered by verticalized LLMs will see immediate results and boost productivity and efficiency at the same time. For example, financial crime investigations often take a long time — manually gathering data, reviewing that data, updating notes on the case file, determining the status of investigations, etc. Gen AI copilots change this dynamic by quickly and efficiently finding relevant data for investigators in a far shorter timeframe than doing this task themselves and highlighting when to add data/evidence to the file to build a complete case file.

Forrester Research notes that ‘Enterprise AI initiative will boost productivity and creative problem solving by 50%’, while McKinsey has written that 60-70% of work that employees currently undertake could be automated with current generative AI technology, with the likes of Microsoft actively investing in the technology. That percentage is surely going to increase as AI application development and LLMs improve, paving the way for broader AI adoption within and across industries at a greater depth than today. Not only will software like the Sensa Copilot be able to help with summarizing investigations and generating narratives, but will undoubtedly also help with other areas such as KYC, CDD, payments fraud, and sanctions screening.

A reappraisal of the use of AI/Machine Learning techniques

The AI machine learning (ML) techniques refined over the last ten years are gaining traction in 2024 due to their predictive abilities. Often forgotten in the shadow of generative AI, predictive AI’s ability to demonstrate risk similarity, abnormal behavior, new patterns, and threats is returning to prominence after being briefly eclipsed by gen AI in 2023.

Three core aspects have changed and are driving this notion:

  • More modern techniques – alongside advancements in the ability of predictive AI, techniques such as deep learning have dramatically improved, facilitating complex pattern recognition and natural language processing tasks with incredible accuracy. Additionally, the notable focus on interpretability and fairness ensures models are not only powerful but also transparent and ethically sound, mitigating risks and biases.
  • Power combo – using either predictive AI or generative AI is nice; however, it’s been proven that coupling these two capabilities in a solution is far more beneficial and complimentary and allows the AI skills to perform better while improving the user experience. For example, modern AI and machine learning techniques help with predictive outcomes in pattern recognition, risk assessment, anomaly detection, etc., and foresee these events and suggest mitigation before they occur. When interlocked with gen AI, the automated content retrieval skill, for example, can use predictive AI output as input/insight to improve the data-gathered picture it provides to end users, delivering a more holistic view of related circumstances and facts while increasing the effectiveness of outcomes.
  • Cost – where enterprises spent an extortionate amount on testing older style AI/machine learning techniques in their environments in the past, the cost of training models has decreased. While the models have gotten better and more complex, we are using larger data sets, so the accuracy has increased. In fact, AI’s accuracy is doubling every six months. As such, it is now more cost-effective to run.
‘One-size-fits-all’ solutions will proliferate in anti-financial crime

To capitalize on the benefits of new tech and AI innovations, a ‘one-size-fits-all’ solution model is gaining traction. The approach of heavy customization or buying or building software that only suits your exact requirements is fading because the cost, effort, and practicalities of continuing this way stifle the budget expenditure and disadvantage the way to consume, upgrade, and modernize operations.

In the case of anti-financial crime, the cost of compliance seems to be rising year-on-year, yet budget approval and spend must now cover more and more every year! A ‘one-size-fits-all’ solution approach can streamline aspects of this – core platform, upgrades, patches, bug fixes, and the ability to implement new AI innovations or update the AI already in place – at a more cost-effective price. This offers greater ROI and bang for your buck in lowering expenditure while optimizing tech capabilities.

As for AI, the four previous trends give this away – investment in AI, the power combo of predictive and generative AI, and the integration of AI for specific use cases, such as in AML compliance, is rising, and adoption rates are growing fast. For ML algorithms, LLMs, and Large Action Models (LAMs) to perform the best or close to the best they can, they should execute fast and integrate with a core platform to deliver results. This is made more difficult when software has been significantly customized or is so bespoke that the ability to upgrade/deploy takes a long time. The process to get it all working is arduous, troublesome, and costly from a resource and budget perspective.  Furthermore, the time it takes to implement bespoke products leaves financial institutions exposed to more security liabilities until their customized product is ready.

By adopting a SaaS product with optimized, consistently competitive performing algorithms for each use-case, financial institutions can dramatically improve the time to value of ML software while also covering themselves from the continuously updated regulatory risks as quickly as possible. Alongside this, the technology and its benefits can advance more quickly as developers can see how the product is being used across the board. The result? Better software, better ROI, up-to-date AI, and a much harder system for financial criminals to exploit.

Greater push for information sharing to combat fraud and money laundering

The need for better data quality and connectivity is a long-standing issue in the anti-financial crime community, though a greater push in 2024 for interconnectivity within an organization, with meaningful external data sources, and amongst peers, is happening amongst C-suite executives as a critical driver to reduce operational burden and impact on staff.

Financial institutions know they have data in other functions of their organization, which could be useful to AML processes and help lessen the friction felt by customers in having to engage with multiple touchpoints. This data, if understood and made available, could help address pain points in understanding customer/risk/changing risk profiles. For example, underwriting and borrowing data from credit functions or other types of non-financial risk data (e.g., legal) could assist corporate clients or provide beneficial ownership details.

Accessing further specialty external data sources that are curated to remove duplicate ‘noise’ or to bring new data fields into the picture, such as ESG data, is also being looked at more closely. Everyone is dealing with the same core challenges but trying different approaches to get to the same end results.

Further, industry reports, such as The Financial Action Task Force (FATF) published report, Illicit Financial Flows from Cyber-Enabled Fraud in November 2023, continue to advocate for better data sharing. In this comprehensive document, one of the main takeaways is the domestic coordination of public and private sectors collaborating via information sharing to mitigate cyber-enabled fraud and money laundering.

It isn’t just a domestic financial services issue either; supporting international collaboration between different industries and sectors will help with investigation and recovering fraudulent proceeds. It is only by working together that companies and their resident countries can react quickly and ideally stop fraud and money laundering before it happens by raising awareness and strengthening their detection processes. This can be achieved through multiple avenues, such as key summits where financial institution anonymity is guaranteed, as well as signing up to and making use of new technologies that seek to address this issue.

Payments are changing globally

Payments will look different in two years’ time than how they look today, fuelled by the expectations around instant payments, ease of cross-border money movements, and changing payment frameworks and regulations.

For example, real-time payments in account-to-account payments (A2A) and greater oversight of digital wallets are on the rise in multiple countries. These include:

  • The ongoing advancement of ISO 20022 and messaging standards
  • PSD3 and PSR1 packages look to “evolve, not revolutionize” payments by removing barriers, making transactions easier, and evelling the playing field for banks and non-banks alike (NB: expect final ruling towards the end of 2024)
  • Further delays in the EU’s SEPA schemes (Sepa Credit Transfer, Sepa Instant Credit Transfer, Sepa Direct Debit Core, SDD Business-to-Business)
  • Ongoing country-level reviews of AML/CFT applicability laws and oversight of digital wallet custodians (fiat and crypto currencies)

These present heightened complexities to adhering quickly, upgrading solutions, and managing new risks.

Alongside this, the US will seek to expand its open banking initiative through schemes such as Pay by Bank. This will ensure real-time money flows and lower costs to merchants, but likely also cause an increase in potential financial crimes.

2023 Trends continuing to be relevant in 2024
AI legislation and AML/Sanctions regulation

Country-specific approaches to regulating AI will continue in 2024, as will principles-based laws that capture AI innovation, AI ethics, and consumer rights. AML/CFT and payments-related legislative changes will also move through parliamentary proceedings across the globe as governments and regulatory bodies take different stances on top priorities.

Therefore, expect to see new proposals, enactments, and implementation challenges for fast adherence, in what seems to be a shorter and shorter timeframe from laws passing to enforcement date.

For example,

  • UK | The Economic Crime and Corporate Transparency Act 2023 (Commencement No. 1) Regulations came into effect on 15 January 2024, impacting aspects of money laundering and fraud, including mixed-property transactions; enhanced due diligence regarding high-risk countries, and information sharing between certain regulated businesses in relation to customers where this may assist in carrying out customer due diligence to prevent or detect economic crime. The Payment Services Regulator’s new directive on 50:50 reimbursement also came into effect, likely costing institutions aligned to fraud/compliance functions, especially where systems and controls are not modernized to spot suspicious patterns and fraud.
  • EU | The European Commission PSD3 proposal (13Jul23), together with a Payment Services Regulation (PSR1) that compliments and clarifies some aspects of PSD3, is expected to be adopted before late 2024 at the earliest, affecting the payments landscape as we know it today.

These changes impact AML, fraud, and cryptocurrency, as well as AI technology innovation. For example, the European Union AI Act is considered the most comprehensive and specific AI-related law to encompass AI development, usage, consumer safety, and privacy, while US President Biden’s Executive Order focus on the safe, secure, and trustworthy development of AI is also expected to reveal its findings before the end of the year. Federal agencies have already taken more than two dozen actions since the Order with more deadlines also following in February, March, and April.

AI-facilitated crime will increase

AI is useful for combatting financial crime, but it isn’t just those on the right side of the law that are using it. AI-facilitated problems are increasing: fake news is spreading across social media, undermining social trust and democracy, while AI-generated images are linked to fake account openings and other persons of interest. In finance, problems primarily center around artificial voice generation or deep fake video technology for identity impersonation, leading to more ways for bad actors to get around security barriers.

A recent UK government report into safety and security risks of generative AI states, ‘By 2025, generative AI is more likely to amplify existing risks than create wholly new ones, but it will increase sharply the speed and scale of some threats. The difficulty of predicting technological advances creates significant potential for technological surprise; additional threats that have not been anticipated will almost certainly emerge.’ As such, 2024 is sure to be fraught with alarming possibilities. Financial institutions will do their best to mitigate such issues by using the latest software that actively combats these problems as they arise.

Money laundering and fraud will increase

With the global economic slowdown of 2023, a cost-of-living crisis caused problems around the world. Today, 1 in 4 people in the developed world are struggling financially, according to the World Economic Forum. As such, it is highly likely that money laundering, fraud, and other financial crimes will increase. Alongside this, more people than ever now have access to AI tools that can be used to commit crimes. Previously, law-abiding citizens may have taken drastic measures to pay bills, with the likes of Authorized Push Payment (APP) fraud spiraling out of control and insurance fraud spiking in the UK. It’s no wonder individuals can fall victim to scams and mule-type activity. According to Zurich, fraudulent property claims increased by 31%.

A rise In inflation has historically been a breeding ground for crime. When inflation decreased in the early nineties, so too did crime rates in the US and Europe. As such, financial crime investigators need to be more vigilant than ever when it comes to risk alerts on people who have never previously been in trouble with the law. Implementing an advanced and robust alerting procedure to catch these instances will help many institutions in 2024.

Increased demand for skilled professionals in a market with a skills shortage

There are many skill shortages in the areas required to support global anti-financial crime efforts. AI use cases are pressuring the demand for skills across roles such as data scientists, sanctions experts, and level 1 and 2 investigators.  Of Forbes’ top 10 most in-demand skills for 2024, five have links to combatting financial crime (gen AI, data, cloud computing, machine learning engineering, and cybersecurity). High demand for employees alongside problematic skill shortages isn’t a recent development either; this has been a growing concern since 2018. Employing those with transferable skills or upskilling those in junior-level positions may be the path for enterprises to keep up with current technology and its future advancements.

The evolving landscape of 2024 and beyond will only add to this pressure and require firms to anticipate changing needs ahead of time while still growing teams to accommodate laws and advanced tech around the world.

Refinement of real-time capabilities to detect financial crime

Following on from previous years, financial institutions are still under pressure from regulators to facilitate real-time transaction monitoring. As digital banking continues to grow – 3.6 billion people are expected to use online banking alone in 2024 – transactions are rapidly growing. Not only are these payments moving between personal and business accounts, but the medium is changing – mobile banking apps, cardless options, digital wallets, etc. Both will continue to disrupt the status quo.

In addition, regulators want financial institutions to provide evidence they are monitoring transactions for AML in real-time; however, specifics on how and what this looks like differs across banks and countries. Banks must decipher how to significantly improve detection monitoring in a way that makes sense without creating an influx of new alerts and missed real-time risk rendering and adhering to existing AML regulations to stop suspicious transactions as they occur. Though this may seem problematic, it also offers opportunity and potential through unified end-to-end systems that can work on all areas of financial crime prevention.

Ongoing complexity and difficulty in implementing sanctions

Thanks to 40 national elections in 2024, many changes to sanctions lists will occur throughout the year, especially for PEPs, spouses/siblings of PEPs, etc. The ongoing conflicts in Iran, Gaza, and Ukraine will also continue to impact sanctions, sanctions evasion, and preventing funding of war activity via trade movements.

AML compliance functions will face these complexities, likely by retraining staff and modern screening technologies (such as watchlist management) where budgets allow, to aid with true real-time capability, more accurate alert generation, and improved case management.

Proactive institutions, using software to combat those on watchlists and trained staff, will find themselves more able to cope with any new sanctions as they emerge. For those that aren’t so ready, implementing sanctions software like SymphonyAI’s Name Screening and Transaction Screening must be a top priority.

Conclusion

The 2024 trends in anti-financial crime are dramatic, somewhat foreseen, but more importantly, solvable with the right know-how, teams, and technology. It is vital to take notice of and take steps towards understanding them in the wider context and within your own organization. It is more important than ever to stay vigilant to the threat of fraud, money laundering, and other forms of financial crime with robust, powerful software such as SymphonyAI – award-winning software used by more than a third of the world’s top 100 banks in their crusade to eliminate financial crime.

A generative AI investigative assistant and the Sensa Investigation Hub can augment your existing legacy software to revolutionize investigations. Book a demo today.

Latest Insights

AI solves sanction evasion
 
06.10.2024 Byline

The threat of sanctions evasion creates challenges similar to those in AML. But AI can help.

Financial Services Square Icon Svg
Improving detection efficiency with AI
 
Video

Improving detection efficiency with AI

Financial Services Square Icon Svg
The benefits of using SaaS in financial crime prevention
 
05.22.2024 Blog

The benefits of using SaaS in financial crime prevention

Financial Services Square Icon Svg