How to work together amus and fraud teams to reduce the risk and to combat financial crime

Fraud and anti-money laundering (financial institution) are teams Finally Start working together! 80% of them indeed. In addition, 53% of the institutions of the AML and financial crime departments share a part or all systems they use, and 100% of the FIS surveyed use artificial intelligence (AI) as part of their arsenal across many different applications.

What's next? Reduction of the total ownership costs (TCO) of systems used by these areas of the institution. Build not only a healthy return on investment (ROI) about such combinations, but also through a “holistic” perspective in AML and fraud data streams and a “broader context for the investigation” in order to benefit the entire institution and its customers.

These are some of the most convincing knowledge in a recently carried out survey of 30 US “MID Market” bank, credit cooperative and Neobank Anti statements and compliance experts, including a handful of IT and operational specialists. The research company Celent and HAWK, a provider of AI supported by AI supported AML and fraud prevention technology, the people surveyed from a mixture of assets, mainly in the range of $ 1 to 20 billion, although smaller and larger institutions were also included.

Don't start technology – start with people and processes

The survey with Hawk's CEO and co-founder of Tobias Schweiger discussed, said the first mistake that many bank innovators and credit cooperatives carry out in the search for AI solutions for AML and fraud prevention and recognition, said with the technology. Instead, he suggested “starting with people and cooperation across processes.

Anyone who works in the financial services industry or has read one of the many articles from FinExtra about the ongoing efforts of financial institutions to prevent, recognize and defeat money laundering – or to have a webinar or take part in a conference session that touches a discipline – you know that this new trend for cooperation is unveiling. This is because many financial institutions – especially those with a larger size – have set up and competent teams and systems for managing KYC and customers transaction -screening in recent decades. They often often indicated in interviews and conversations that not many of their results or data between the departments.

Even with sophisticated programs and systems to bring the right customers on board and protect themselves against the fraudsters – always in the attack to steal the money of their customers or to damage the reputation of their institutions, or both – only a few institutions in every size, until recently, have made it easy or even mandatory that they regularly communicate for their internal fraud and aml teams and help with their counterparts with their counterparts convey. This finally changes according to the answers that were shared from 2025 Trends in the AML & fraud convergence at the US bank survey.

Not always easy, but the survey says that the cooperation between AML and anti-frago teams works

Do not call this new collaborative discipline “Framl” because 67% of those surveyed have raised objections to the term and explains that he “corresponds to the complex processes” in order to align fraud and aml team efforts in their institutions. Schweiger also pointed out that KYC and extended Due Diligence (EDD) functions selected by banks are the most common consolidation areas that are case management and transaction monitoring are now also part of the mix.

“The combination of processes enables information to flow freely,” he said between the functional areas of the back office. The “tech” is now better and many institutions now have comparable systems to master both AML and fraud.

Many larger institutions have known and proven for decades how helpful mechanical learning (ML) can be when examining large data records in order to identify suspicious activities and unknown transaction patterns. ML scripts and processes have been implemented to automatically reduce the risk of human errors and to perform repeating tasks more efficiently than live employees in many back office departments. Schweiger says that the transition from reactive, inventing recognition to proactive, predictive analysis and use of the uniform data and a larger context that a combined or “framl” approach can provide can determine AI risks more effectively and reduce incorrectly positive activities.

Banks who were asked by Celent and HAWK evaluate concerns about compliance with compliance, the arising types of financial crime and the increase in regulatory expectations in their answers to a question about the “top drivers of change” in their fraud and aml programs. However, the customer's friction kept as low as possible and continuously to support digital financial services for their customers (develop further).

Top challenges for smaller FIS: Finding, keeping employees and reducing wrong positive aspects

The biggest challenges – according to the survey – do you do all of this? “Analyst personnel” and high rates of false positive warnings. The latter occur when the names, locations or transactions of people trigger or apply alarms in processing because they are classified as incorrect or fraudulent. Then these warnings prove to be incorrect – they consume valuable time for compliance employees and can make overloaded teams and the need to make more available to “get ready with the workload”.

According to Schweiger, the Cross training of system users becomes more important for all institutions in both fraud and AML teams. He also noted that AI can be of great help in solving the concerns and burden on the staff and at the same time reduces false positive results:

“AI is simply more precise when dealing with the complexity of cases,” he said, noting that the modern software for biometric behavior biometry goes far beyond earlier test. “Generative AI has the ability to look back historically in transactions, segments and behaviors (the actors involved), and deviations can be clearly seen,” said Schweiger.

By using AI to first define normally and then proceed from this point, banking systems can find anomalies, reduce false warnings and immerse themselves in certain data elements, which, for example, are not caught by a “hard and fast” rule of screening.

AI is used to present data cleaning, speed examinations and the simplification of the reporting

After reducing false positive things, the AI ​​plays the most frequently in the data cleaning, the tightening of examinations and the automation of the ad hoc website.

It also helped 27% of banks to write their suspicious activities reports (SARS) via certain questionable transactions. Many other useful and related fraud/AML applications are listed by more than 25 to 30% of the institutions surveyed in the survey. It is therefore clear that the KI/ML dynamics is strong and will probably continue to become stronger if forces between protective departments and systems are connected.

Financial effects of AML/Anti-Fraud consolidation can be extremely positive for institutions

There is great financial incentive to combine forces. Banks surveyed said they had saved money from fraud and aml cooperation as well as the consolidation of efforts and cooperation with fraud and AML saving or already impressive savings. The figures cited were between hundreds of thousands of millions of dollars a year in operating and technical costs. In fact, two thirds of the respondents who have already carried out that the consolidation of their AML and fraud systems have saved them at least $ 1 million a year, and about 50% say that their expenses have dropped by more than $ 5 million a year.

The survey finds that sometimes the combination and cooperation between fraud and Kyc/Aml teams are unsuccessful, and while budget, personnel, training and legacy technology are obstacles to effective consolidation programs, the “cultural changes” and differences in the focus and specific activities that prevent everything from working together, especially in the case in which there are some earlier Departments, the departments, the results of the requirements.

Talks begin best to find the best approach and technology for the needs of the institution

Schweiger confirmed that both best practices and the survey results show the intelligent path to ensure the cooperation and consolidation between anti-woman and AML functions that achieve the goals desired by financial institutions to begin discussions across departments, systems and processes. Since he explained: “If you do not do this, you do not know which technology change you want. People and processes first, then the ideal approach for banks and credit cooperatives that find it in the second step is to find a targeted operating model (TOM).”

While the institutions of the technology that they have and that have recently been available to their specific needs, he pointed out that some banks and credit cooperatives have found it useful to start with smaller pilot projects within AML and/or fraud default departments. The “analytical work” carried out in these first studies can help to provide cost savings and process improvements from the complete implementation of the proposed cooperation plans.

If Kyc, AML and Anti-Frags employees and systems do not work more active, new technologies and innovations such as AI, mechanical learning and data identify their combined functions more effectively and more securely, there will continue to be a lot of wasted time and wrong notifications about transaction screening, on board and other areas of exams, said Schweger.

In the meantime, the criminals will continue to meet financial institutions and their customers from all sides, regardless of which area or a system is trying to prevent them from it or even catch them on the plot. It would be a real shame for the entire institution and its customers, he said, since “they will miss the financial crime by not working together-and vice versa were not taken into account.”