This is a guest post by Barley Laing, the Managing Director at Melissa
AI-driven technologies are driving fraud worldwide. AI-powered deepfake software, automated phishing systems, synthetic identity fraud, chatbots and advanced data analytics are enabling scammers to create highly personalised and convincing fraudulent schemes, quickly and easily.
This growth in AI-powered fraud is backed up by recent research. According to UK Finance in the first half of 2025 there were over two million confirmed cases of fraud, a 17 per cent increase on this time last year. Also, a recent report by Cifas and Experian highlights that identity driven fraud has risen by 9.2 per cent in the last three years, with AI generated identities now making up 42 per cent of all cases in the UK. The same report identified potentially fraudulent savings account openings rose by 53 per cent over the last three years, largely driven by criminals creating identities using AI.
Furthermore, Deloitte predicts that generative AI could enable fraud losses in the US to reach $40 billion by 2027, up from £12.4 billion in 2023.
The big question is how to prevent AI-powered fraud in 2026, to protect customers, your bottom line and reputation?
Collect accurate customer contact data
A good place to begin is by ensuring the accuracy of customer contact data; which is a basic step in the know your customer (KYC) compliance process that a number of financial institutions still don’t always deliver on. Access to accurate customer contact information, such as name, address, email and phone number, makes the verification process more reliable. Having quality customer data enables ID verification technology to confidently cross-reference the provided information against official databases, or other authoritative sources, without inconsistencies that could lead to false positives or negatives. Additionally, with fraudsters exploiting inaccuracies in contact data to create false identities and manipulate existing ones, discrepancies in a user’s phone or email, or an address linked to multiple identities, could be a red flag for extra scrutiny.
In fact, address verification is recognised as the cornerstone of contact data quality, because having access to up-to-date customer addresses makes it much easier to match and verify identities across multiple sources. This means verifying the accuracy and legitimacy of an individual’s address should take place at the beginning of any identity related process.
Lookup / autocomplete technology
Having access to lookup or autocomplete technology is a good way to obtain accurate customer contact data. With an address lookup tool, for example, it’s possible to deliver accurate address data in real-time by providing an appropriately formatted, correct address at the onboarding stage, when the user starts to input theirs. These services are vital when around 20 per cent of addresses entered online contain errors; these include spelling mistakes, wrong house numbers, and incorrect postcodes, due to errors when typing contact information. These tools also significantly streamline address entry, reducing keystrokes by up to 81 per cent. This speeds up the onboarding process and improves the whole customer experience.
Bear in mind that similar technology can also be used to verify email and phone details at the first point of contact, ensuring these valuable contact datasets are verified in real time.
Deceased and change of address data
Synthetic identity fraud, where criminals combine real and fake personal data often using AI tools to create a new identity, along with some other types of fraud, usually rely on outdated, inactive, or deceased identities. This makes scanning for deceased and change of address details important, particularly as these are areas that traditional fraud controls may overlook.
For best results incorporate National Change of Address (NCOA) data from the likes of the Royal Mail or the United States Postal Service (USPS), along with deceased suppression data, into identity verification and customer onboarding workflows. This way financial institutions can proactively shut down key entry points for fraudulent activity.
Fraud prevention with electronic identity verification (eIDV)
Data hygiene practices alone aren’t sufficient for effective KYC or fraud prevention, which is why using an electronic identity verification (eIDV) platform is vital. These ‘always on’ tools can, in real-time, cross-check the names, addresses, email addresses and the phone numbers provided by prospective customers when remote onboarding – providing a good customer experience, while helping to prevent fraud. When compared to manual checks using such a service is significantly quicker, more accurate and cost-effective way to implement ID verification and prevent AI-driven fraud. Additionally, the technology doesn’t require additional staffing or training costs, and using it minimises the risk of human error.
Match IP address to home address
Finally, consider using geolocation technology as part of the identity verification process to outsmart fraudsters using AI. It can validate whether an applicant’s physical location, via their IP address, corresponds with the address and other personal details they provide. For example, if an applicant for a service or product claims to reside in London, but the IP address data from their mobile device reveals they are actually 7,000 miles away, this discrepancy serves as a big red flag. Therefore, undertaking an additional verification procedure would be wise prior it being approved.
To sum up
With AI starting to supercharge fraud and scams in 2026, those in financial services need to be on the front foot to prevent it. What is needed is a focus on data quality and eIDV technology as criminals increasingly use the fast evolving technology at their disposal.
Image provided by Melissa
