DoorID Launches To Take On AI Identity Fraud At The Front Door

DoorID launches to help lenders and other businesses check that a customer can physically access the address they have given, at the moment it matters most: before money or goods change hands.

Founded by Paul Williams with his brother and co-founder, Iain Williams, DoorID uses a guided smartphone check to connect a person to a property. It adds live evidence of physical access alongside existing identity and fraud checks.

Customers receive a link and complete a guided check via their own phone’s browser. The check takes about 70 seconds, with no app to download. The check brings together identity, location presence, capture integrity and property evidence. On completion, DoorID returns a ‘Pass’, ‘Refer’ or ‘Fail’ result with a tamper-evident evidence pack.

Businesses choose where the check sits in their onboarding journey, such as before releasing funds. Verification data is handled in line with UK GDPR.

The idea for DoorID came from an unexpected source. A genuine customer of Quick Car Finance, the brothers’ Bournemouth-based car finance brokerage, which employs around 60 people, was away on holiday when asked for a photo check. Instead, she sent an AI-generated image of an open front door. It looked real, but Paul spotted that something was off.

Paul Williams, founder of DoorID, said:

“She was a genuine customer who simply wasn’t at home, and she’d used AI to make a picture of an open door. If someone would do that for convenience, we had to ask what a determined fraudster would do.”

He added:

“Documents can carry the right name and address without proving that the person can get through the front door. The answer isn’t more hassle for every customer or more cumbersome onboarding. It’s intelligent friction: one quick check at the moment the risk is real, which protects genuine customers as much as the business.”

The launch follows a demonstration in which the team used a freely available AI tool to change the photograph on a driving licence image in 28 seconds. The exercise illustrates how easily AI can manipulate the visual evidence in an identity document. It was not a test of whether the altered image would pass a lender’s checks.

During the first half of 2026, Cifas recorded more than 220,000 fraud-risk cases on the UK National Fraud Database, its highest January–June total. Identity fraud accounted for 59% of cases, rising 9% year on year to nearly 130,000. Impersonation using a victim’s current address remained the leading reason for identity-fraud filings and increased by 12% compared with the same period in 2025.

The brothers’ experience comes from running Quick Car Finance, which they started after helping their mother find a car and arrange finance. The brokerage has been trading for nine years. Paul built DoorID, drawing on that first-hand knowledge of customer journeys and fraud risk.

DoorID has been tested through Quick Car Finance. It is designed for any business that needs to know a customer can really get to the address they’ve given, including banks, mortgage lenders, conveyancers, asset and device finance providers, telecoms providers, insurers and housing providers.

Businesses can request a DoorID demonstration here.