Can Facial Recognition Stop Shoplifting? A Look at the Tech & Potential Concerns

A shocking number of 8 million shoplifting cases were recorded in 2022-2023 in the UK. It cost the retailers nearly £1 billion (US$1.2 billion) in just 12 months. To minimize shoplifting cases, the UK’s major retail representative authorities and police announced the Retail Crime Action Action Plan in October 2023 that invited retail store owners to submit CCTV footage of shoplifting incidents taking place at their stores.

UK’s police intend to use this footage through a digital imaging system and employ facial recognition technology to identify shoplifters. This anti-shoplifting movement also includes a distinct program known as ‘Project Pegasus’ that enables comprehensive intelligence to identify and collect information on organized crime associations across the UK. Pegasus is a unique information-sharing platform between the retail industry and policing authorities.

We witness the use of facial recognition technology being used through CCTVs in big retail stores. But the real challenge occurs for facial recognition technology when shoplifters use bypassing tactics to spoof facial recognition and steal away precious merchandise causing heavy losses. 

Key Takeaways:

• Shoplifting is a major concern, with the UK alone reporting 8 million cases and incurring losses of around £1 billion annually.
• Traditional methods have limitations. EAS tags, for instance, only detect unpaid items at exits, not the shoplifters themselves.
• High-resolution cameras equipped with accurate facial recognition software can identify shoplifters from a database, potentially preventing crimes before they occur.
• Fast identification is crucial to apprehend shoplifters before they leave the store.
• Solutions like Facia offer high-speed verification (under 1 second) to identify potential shoplifters.

Let’s find out how shoplifting can be curbed through advanced facial recognition technology.

What is Shoplifting

Shoplifting is the act of stealing merchandise from a retail store. Shoplifters enter the stores usually in disguise and then pick the items they want to steal and hide them in their purses, bags or even garments they wear. Sometimes, shoplifting is also referred to as casual theft. Shoplifters mask this activity by purchasing other items that are less costly to minimize suspicion. Shoplifting is a crime in every jurisdiction. Mainly in the UK and the US, it is a punishable crime where the defaulter can be fined, forced to community service, apy the damages to the store owner or even imprisoned if the act of shoplifting is repeated.

How Shoplifting can be Reduced at Retail Stores?

Shoplifting is a tricky crime and a high level of alertness is required to timely catch a shop thief. However, with the growing retail industry, stores are expanding their physical location as well as sizes with the rising demand. So, hiring a separate workforce and checking each and every customer with suspicion will not only bring in hefty costs but also cause a loss of customer trust and convenience while shopping. 

There are two major ways to identify and detect shoplifters:

EAS (Electronic Article Surveillance) 

EAS is a security surveillance technique which is commonly used at retail stores and hypermarkets. It includes an EAS Antenna gate installed at all the exit points of the store. Every product inside the store has an EAS Tag on it. If a customer has picked the product and not paid for it on the counter, the system automatically detects this product once the customer reaches the exit gate where the EAS antenna detects an unpaid product tag and raises an alarm.

Due to the technological limitations, this technology seems insufficient in fully tackling the shoplifting issue.

The main challenge with EAS tags is that the shoplifter has already entered the premises and lifted merchandise and now there is a very limited time and window in which the system has to catch the criminal. The only way left is to detect the product’s EAS or call the store’s security or police to catch the criminal.

The law enforcement will also require CCTV footage and other information as a proof to identify the criminal.

Facial Recognition Technology

Facial Identity Verification solution providers can play a vital role in the detection and mitigation of shoplifting. CCTV cameras are the best ways to collect customer’s facial data and match it with a criminal’s database. For this purpose, the facial recognition technology should be accurate enough to minimize the errors of False Match and False Non-match Rates in the identity verification process.

There are other considerations in this regard such as:

• The distance and angle of CCTV cameras in accurately detecting facial identity features impact the overall facial recognition procedure. Therefore high-resolution 2K cameras are preferred with fine zoom-in capabilities in order to record clear videos of faces.
• Liveness detection is another important feature that should be incorporated to detect anomalies and identity theft attacks like Mask Attack.
• Speed in matching biometric facial data with that of criminal identity accurately is one of the key factors in detecting shoplifting through facial IDV solution.

How Facial Identity Verification Solutions Can Enhance Detection of Shoplifting?

Firstly, IDV Solution providers, their business clients and the end customers need to realize the importance of facial recognition as a core component to prevent Shoplifting. It is not only that the shoplifters pose a threat to the retail store owner only. They can pickpocket any other customer or even pin a shoplifting activity onto a genuine customer or even steal their credit card to make payments. Things go deep when the element of mask attacks, deepfakes and identity theft is combined with this malicious act.

So, firstly, facial recognition software needs to understand all possible threat vectors that support shoplifting. This includes the use of Artificial Intelligence (AI) in spoofing identities, designing 3D masks, helping in bypassing CCTV footage or injection attacks in CCTV camera systems at the retail stores and supermarkets.

• Facial recognition solutions need to upgrade and integrate AI for detection of AI-supported shoplifting attacks.
• Face verification solutions also need to incorporate smart systems to differentiate between a video injection and a real CCTV footage in a timely manner.
• They also need to consider the cost element where small retail stores can also enjoy a high level of security against shoplifting. IDV solution providers can cut costs by employing AI and other such mechanisms that can help small retailers in preventing shoplifting as well.

Facial IDV Solutions also require high speed in detecting shoplifters as they have a very small window of time in which a thief enters and leaves the store. Therefore, retailers can opt for Facia which is the world’s fastest facial verification solution with speed of less than 1 second.

Facia is a cutting edge Identity Verification solution suite that is tested by NIST for its performance and capabilities of identifying and verifying facial identity anywhere in the world. It uses both 1:1 and 1:N verification models suggested by NIST to accurately and swiftly identify and differentiate between a genuine customer and a criminal.

ISO 30107-3 Level-1 and ISO 30107-3 Level-2 compliance, it comes at the forefront of fighting against illicit crimes like shoplifting and helps retailers maintain their store’s security and integrity in the fastest way possible.