How to prevent AI-generated fake customer accounts

Evolving technology and increasing digitalisation are facilitating scammers in creating fake accounts for committing fraud.

It’s why fraud is on the increase with Cifas, the UK fraud prevention service, revealing in its annual Fraudscape report in 2025 that 421,000 fraud cases were filed to the National Fraud Database (NFD) in 2024 – a 13% increase on 2023, and the highest number on record. Identity fraud remains the most prevalent type with 250,000 identity fraud filings recorded in 2024 – an increase of 5% on 2023 - accounting for 59% of all filed cases to the NFD.

The implementation of effective customer verification processes is essential because of the role they play in identifying and preventing the creation of fraudulent accounts.

With such procedures in place you can not only verify the name of a customer, their contact telephone number, postal address, and email address, but also source their device location in seconds. You then have certainty that the customer is who they say they are.

The four key steps for customer verification:

1. Integrity verification: For confidence that you have the correct details to begin with check every core data point the prospective customer provides - postal address, phone number, etc - against dependable global sources of data in real time.

2. Augmentation and extension: When scanning against authoritative global sources of data from government, utility and credit agencies, augment customer records with missing contact information, which further supports customer verification. For example, a missing postcode can be added to an address, a phone number, email ID and its status to a customer record. This process also aids organisations in improving customer communications, enabling them to deliver a standout user experience and drive revenue.

3. Consistency check: This step connects the core data points with the information sourced in the action above. The cross-checking between the two sets of data helps to determine consistency, and supports the standardisation of contact information formats, making it easier to manage and analyse.

4. Quality Assessment: The last action focuses on the overall precision and quality of customer records. This involves every element of the record receiving a quality score that is numeric, with the record itself assigned a confidence score. With this insight spotting and eliminating customer accounts that seem high-risk or suspicious is expedited.

How can these steps be effectively delivered?

Undertake name verification at the customer onboarding stage with technology that can analyse, format and gauge name accuracy. These tools must highlight names that are fake or bogus, with verification filters used for the purpose including nicknames, vulgar words, and celebrity names.

Always match a name to an accurate, verified address, because any discrepancy could highlight a possible fraudster and must be investigated further. In fact, implementing address verification to deliver address accuracy is very important because inaccurate addresses can be costly to your business when it comes to shipping, with returns and redeliveries, and the negative experience customers receive because of any delays.

To prevent this it’s vital to use an address lookup or autocomplete tool that provides an accurate, properly formatted, standardised address at the customer onboarding stage as the user starts to enter theirs. Additionally, similar tools can correctly collect email addresses, telephone numbers and names at the first point of contact. These all play an essential role in supporting customer verification.

Sourcing the prospective customer’s location via their IP address is a critical part of the customer verification process. IP addresses must be cross-referenced against the addresses provided for shipping and billing, to detect transactions that are risky or fraudulent. For example, further checks are necessary if the IP location of the customer’s mobile device is nowhere near the shipping address they have supplied. 

In summary

By implementing processes that verify contact details, including name, address, email ID, contact number and IP address, it’s possible to ensure the quality and accuracy of a customer account. This provides effective customer verification, while avoiding AI-created fake records. It’s an approach that not only prevents fraud but strengthens marketing strategies by supporting informed targeting.

This guest blog was written by Barley Laing, the UK Managing Director at Melissa. To learn more about Melissa please visit their LinkedIn page.