Who is faster: a cybercriminal or the police?
Be it phishing emails or helpdesk employees who ask you to transfer money— digital crime is on the rise and is becoming even slicker. Artificial intelligence can play both a negative and positive role in this. On the one hand, new forms of cybercrime are emerging from AI; on the other hand, AI is helping to trace criminals. We spoke to Prof. Frank Thuijsman and Prof. Anna Wilbik, both professors of data science, and Bart Seuren, coordinator of the Limburg Police Department’s cybercrime team. Thuijsman and Wilbik joined hands with the National Police Force to develop new tracing methods. What are the challenges faced in tracing digital crime? What opportunities do new (AI) technologies offer?
Using data and applied mathematics in tracing
Knowledge and techniques from applied mathematics, AI and data science can help us to trace cybercrime more efficiently. Both the police and the UM research team are convinced of this. It was Frank Thuijsman, Professor of Strategic Optimisation and Data Science, who came up with the initiative for the collaboration. He describes: “The police wanted to do more with smart technologies to trace cybercrime. They had already published a job vacancy to this end. My first thought was, ‘Could we not turn this into a research project?’. Then we would be bringing the knowledge of multiple PhD students and their supervisors together. And that’s what happened.”
Thuijsman explains: “We want to get a handle on all forms of cybercrime and be one step ahead of cybercriminals by better recognising their behaviour through data and AI. To this end, we research which possibilities data and AI offer. What works well and what does not? In which directions could we develop even further?” His colleague, Anna Wilbik, Professor of Data Fusion and Intelligent Interaction, adds: “To trace crime, you often have to collate all sorts of data types: social media messages, telephone data, GPS data, etc. That is a challenge in itself. Through our research, we are helping the police to combine data from all sorts of sources and to give meaning to this.”
Forms of digital crime
Digital crime takes all sorts of forms. You probably already know about phishing emails, but that is not the only form. Bart Seuren, coordinator of the Limburg Police Department’s cybercrime team, explains: “We also see, for instance, a lot of bank helpdesk fraud and the hacking of online accounts or of crypto wallets. In addition, more advanced forms of cybercrime exist. Like ransomware or CEO fraud.”
Wilbik adds: “Nowadays, young people are getting increasingly affected by cybercrime. This is apparent from an NOS Stories video: of all the 1,329 cybercrime cases that were reported in 2022, 47% involved someone under the age of 21.” She continues: “TikTok is full of all sorts of videos like this, in which young people get hooked by messages on getting rich quick. They then make their bank card available for this sort of fraud, for instance. Often, it is only the people who take part in these schemes that get caught—the originators behind them are harder to trace.”
Prof. Frank Thuijsman is Professor of Strategic Optimisation and Data Science. He has worked at the Department of Advanced Computing Sciences (and its predecessors) of the Faculty of Science and Engineering since 1987. In addition, he is a member of the management teams of the Department of Advanced Computing Sciences, of the Mathematics Centre Maastricht and of Regitel. He was also recently appointed Secretary of the Game Theory Society. Furthermore, he is the initiator of the KE@Work Honours Programme. Thuijsman obtained his PhD in mathematic game theory at UM in 1989. He has briefly worked for UCLA and The Hebrew University of Jerusalem.
Cat-and-mouse game
The most important challenge in tracing cybercrime is that digital criminals are constantly coming up with more forms of crime. Seuren compares this to a game of cat-and-mouse. “On the one hand, we are constantly developing new tracing methods, partly thanks to AI. On the other hand, criminals are also using new AI techniques. As soon as we develop new tracing methods, criminals come up with new forms of cybercrime.”
Wilbik explains that they also take this cat-and-mouse game into account in their research. “If criminals know what you are working on, they will adjust accordingly. Even during the preparation phase of new tracing methods, we consider how criminals will respond to it and we also adjust accordingly.”
Digital crime is a game of cat-and-mouse. We increasingly develop tracing methods, while criminals increasingly develop new forms of cybercrime.
Other challenges in tracing cybercrime
In addition to the cat-and-mouse game, Seuren also identifies other challenges in tracing cybercrime: “An extra challenge in tracing cybercriminals involved in very common forms of cybercrime is that this type of crime is scalable. It is relatively simple for one group of culprits to affect a great many victims throughout the entire country. This generates multiple cases in one go, and often across different police regions.”
Seuren has also noticed that cybercriminals are using an increasing number of techniques in the hope of not getting caught. “Cybercriminals try to protect themselves well by, for instance, using VPN connections or proxy servers (ed.: a computer located between the user and the internet). It is a challenge for us to break through these techniques in order to find the culprit. In addition, cybercrime does not take national borders into account, while international tracing does. For this reason, lengthy legal aid processes are often required to safeguard tracing on an international level.” “But”, Seuren adds, “by pursuing national and international collaborations with other police departments and with public and private parties, like UM, we can increase our strength.”
The role of artificial intelligence
Artificial intelligence plays an increasingly major role in digital crime. On the one hand, it offers opportunities for the police to trace cybercrime. On the other hand, criminals use it to devise and carry out offences. Seuren elaborates: “We see AI as an extra factor to take into account in tracing investigations. AI can help cybercriminals to make their deceptive activities more believable. Through photo editing or cloning someone’s voice, for example. Investigating the authenticity of evidence material therefore becomes an even more important topic in tracing this crime.”
But Seuren remains optimistic: “The police is becoming more familiar with digital tracing methods and cybercrime cases are being handled more broadly in the organisation. Clustering these cases makes it more efficient to handle them. Even the new digital technologies leave their own traces, so there are always possibilities for the police to tackle cybercrime.” For instance, there are possibilities to link AI to current tedious or complex analysis techniques. Seuren: “AI can be deployed, for instance, to sharpen poor-quality camera images, through which suspects become more recognisable. And AI was successfully used in a cold case to allow a deceased victim to make a call to action.”
Prof. Anna Wiblik is Professor of Data Fusion and Intelligent Interaction at the Department of Advanced Computing Sciences of the Faculty of Science and Engineering. She researchers how we can collate the value of data from various sources and of different natures (text, numbers and images), on the basis of which we can take decisions. Wilbik obtained her PhD in Computing Science in Warsaw in 2010. Thereafter, she worked as a postdoctoral fellow at the University of Missouri. She was also assistant professor in the Information Systems Group of the Eindhoven University of Technology. Furthermore, she is actively involved with the IEEE Computational Intelligence Society.
Pattern recognition
Thuijsman and Wilbik also envision an important role for AI in tracing cybercrime. Thuijsman explains: “Through AI, we can recognise even more patterns and, as a result, filter out cybercrime. For instance, AI can recognise people based on certain spelling errors or the speed at which they type. If there is suddenly a major discrepancy in this, then you can tell that something may not be right. Also think about the words that people often use. If a filler word that a person usually does not use is suddenly often present in a message, this can also be a signal for spam filters.”
Amused, Thuijsman adds: “Everyone has their own way of ending a message, for instance. If my children, who are not from Brabant, suddenly wrote ‘Houdoe’ at the bottom of their emails, then I would know that something was off. These are patterns that you can also teach AI tools.”
Thuijsman concludes: “Back then, we learned not to open the door to strangers. Now, something similar applies, but in a digital manner. Be careful if a stranger calls you or sends you a message. It is better to delete something and make the person send it again than to fall in the trap.”
Text: Romy Veul
How can you avoid cybercrime?
What can you do yourself to avoid becoming a victim of cybercrime? According to Seuren, these are the most important tips (see also the police’s tips):
- Update your software regularly.
- Use a virus scanner.
- Set different, long passwords (and use a password manager).
- Use two-step verification when logging in.
- Guard your personal data well.
- Always consider links critically before clicking on them.
- Know who you are chatting with.
- Make back-ups of your files.
- And if you suspect cybercrime, report it.
Also read
-
Lee Bouwman, a vascular surgeon and endowed professor of Clinical Engineering, specialises in the implementation of groundbreaking healthcare technologies. The key to success, he says, lies in the collaboration between engineers and clinicians. This approach has already resulted in a range of...
-
On April 19, during her inaugural lecture, Anna Wilbik explained how we can squeeze out the whole potential of data to the last drop.
-
Adriana Iamnitchi’s career appeared like a bed of roses. She served as a full professor in the Department of Computer Science and Engineering at the University of South Florida. Nonetheless, she sought out new challenges and moved to Maastricht, where she focused her research on social media.
