The technological revolution we are currently experiencing has brought many advantages to both businesses and individuals. However, where technology has encroached into every single part of our lives, cybercrime has quickly followed. It’s a major problem for the financial services industry and many financial institutions have turned to AI to tackle it.
Cybercrime – a problem too big for humans
The term cybercrime encompasses such a broad range of criminal activity that it is almost impossible to put a figure on the global cost each year. It includes credit card fraud, malware and ransomware, data breaches, identity theft, cyber extortion, crypto-jacking, cyber espionage… the list goes on. Suffice to say that cybercrime costs individuals and businesses around the world billions of dollars every year.
The more technology we have, the more points of vulnerability or attack vectors there are for cybercriminals to exploit. This ever-growing attack surface requires constant vigilance which is simply beyond the capability of us mere mortals.
But AI does a much better job. Predictive artificial intelligence can monitor for security breaches much more effectively than a human can.
What exactly is Artificial Intelligence?
Any technology capable of copying the human decision-making process to perform tasks using computer science is categorised as Artificial Intelligence.
If you have conversed with a chatbot, watched a film recommended by Netflix or opened your iPhone using ID, you have AI to thank. Those suggested quick replies in gmail: AI. That Uber you took yesterday: AI. You get the idea. AI is literally everywhere.
In the field of cybersecurity, AI has evolved through three ‘waves’ to reach the extremely advanced levels of human intelligence it displays today. In the early days, programmers wrote code and AI followed the code, overseen by the programmers. But the programmers are becoming obsolete. Now, predictive AI is not only able to acquire, store, process and apply previous knowledge, it can also analyse data, come to its own conclusions and learn from this process.
AI’s self-learning capabilities make it an extremely powerful tool in fighting cybercrime.
Five ways AI is fighting cybercrime in the financial services industry
Here are five examples of how AI-powered automated fraud prevention tools are being used by banks and other financial institutions to tighten their defences in the fight against cybercrime.
1. To recognise fraudulent documents
Financial institutions must adhere to Know Your Customer regulations to avoid fines to comply with regulations in many jurisdictions. This means proving a client is who they say they are by verifying documents. AI can be used to detect fraudulent documents such as bank statements, invoices and payslips submitted during the onboarding process. It does this by looking at patterns of behaviour and spotting anomalies such as logging in from different devices and locations. These are then flagged for further investigation.
2. To highlight potential fraudulent behaviour
AI is being used to analyse behavioural patterns to ascertain baselines for what is considered ‘normal’ behaviour when filling in forms. It looks at variables such as typing speed, hesitancy, keystrokes, and form corrections. When a form is filled in, if behaviour deviates too far from the baseline, for example, things like name and address are being copied and pasted or the form filler is hesitating over information that they should know, AI can highlight the unusual behaviour so that it can be analysed further.
3. To fight mobile fraud
Near-field communication (NFC) technology enables us to pay for purchases using our smartphones. While certainly convenient, it does leave all of us open to hacking and cyber threats. AI can detect abnormalities based on customary user behaviour and trigger additional verification to minimise the risk of fraud.
4. To combat email phishing scams
Phishing scams involve criminals posing as a trustworthy source to persuade individuals to hand over passwords and other personal information. They are becoming ever more elaborate and cost individuals and businesses millions every year. Financial services companies are investing in AI-driven email protection systems to identify fake messages and links. Automated methods of detection based on machine learning analyse and categorise mail extremely fast to help eliminate spam and the numerous types of phishing scams.
5. To identify insurance fraud
According to the Association of British Insurers, in 2020 insurers detected 96,000 dishonest insurance claims valued at £1.1bn. It’s no surprise that the industry has invested huge sums to find AI solutions to reduce fraud. Machine learning’s superior pattern recognition capabilities speed up the claims validation process and can accurately identify fake claims.
I have over 20 years of experience in the financial services industry and hold a Chartered FCSI qualification. I ensure that our operations are fully compliant with the rules of our most stringent regulators.