Stopping loss at the checkout is one of the greatest challenges retailers face today. And while self-checkout kiosks are easy targets to blame for the bulk of this loss, manned checkouts also present plenty of opportunities for theft.
“Shoppers are learning the constraints of the checkout process, and then they can exploit the process and technical shortcomings to their advantage,” said Zebra Technologies Senior Solution Manager Andy Doorty.
“Retailers are struggling with losses at the checkout due to a mix of challenges—employees giving unauthorized discounts, customers skipping scans at the self-checkout, and fraud slipping through unnoticed,” added i3 International Chief Customer Officer Vy Hoang. “With fewer staff on the floor and rising organized retail crime, it’s becoming harder to catch mistakes and intentional theft before they impact the bottom line.”
Thankfully, recent advances in artificial intelligence are proving to
be a game-changer in helping retailers stop loss at the checkout.
Checkout Challenges
Retailers are in a sticky position—not only must they work to stop loss at the checkout, but they also have to keep the customer experience top-of-mind so consumers continue patronizing their stores.
“Checkout loss management is a balancing game between reducing loss while providing the best customer experience possible—the more friction you introduce, the unhappier your customers become,” said Dragonfruit AI Cofounder and CEO Amit Kumar. “So, any efforts to reduce loss must be balanced by the amount of friction they cause in the transaction and provide a clear ROI.”
Finding that balance is much easier said than done.
“Retailers are trying to find the right balance of convenience, speed, and cost,” said Zebra Global Retail Technology Strategist Matthew Guiste. “Self-checkout improves some costs, like lower labor, but can increase costs, like loss.”
Scan avoidance, ticket switching, and faking the payment process are all ways in which theft can occur at the self-checkout. But, with consumers growing more comfortable with the self-checkout experience, Guiste says it is imperative for retailers to figure out how to close down these theft opportunities.
“As retail continues to evolve, checkout losses will likely become even more complex with the rise of AI-driven fraud, more sophisticated organized retail crime, and increased reliance on self-checkouts,” Hoang added. “Retailers will need smarter solutions that go beyond traditional loss prevention methods.”
Internal theft can also occur at the checkout.
“A foundational piece of retail businesses is the employees,” said Solink Senior Leader of Asset Protection Cathy Langley, LPC. “As most of us already know, high turnover rates of store employees, including cashiers and assistant store managers, create more potential for cashier thefts, discount abuse, fraudulent transactions, and more. The core challenge for retailers here is how to retain employees or train and monitor new hires quickly.”
Langley added that retailers also face the challenge of auditing their loss and being able to provide evidence of that loss: “With separate data and video systems, or camera infrastructure that is not properly positioned to collect shopper insights in the self-checkout area, retailers will find it challenging to pinpoint where loss is happening and understand effective solutions. When pilot programs for potential solutions are conducted, retailers will also face the difficulty of proving the ROI because they won’t have the tools at hand to accurately track and monitor checkout area success.”
How AI Can Help
For years, companies have been promising artificial intelligence would help solve all our LP problems—and we’ve been left waiting. But now, the technology has finally caught up, and it truly can make a huge impact on retailers’ bottom lines.
“Until very recently, there had really been very few solutions for helping retailers manage checkout loss,” Kumar said. “With today’s video AI technology, there are now new approaches with a much lower total cost of ownership that enable retailers to tackle these challenges. These new approaches can help retailers detect and deter loss at SCO as well as for staffed lanes and even across the entire store.”
Langley agrees, saying that AI has already started to transform how retailers look at loss.
“The retail industry has already witnessed the shift from reactionary measures like reviewing video streams or POS data to track theft post-event, to preventive measures like auditing stores in real-time to ensure that the staff are actually at the till when customers are present,” Langley said. “AI has been a huge help with driving this shift, and it will be even more useful when encouraging retailers to now take on predictive measures rather than relying on preventive measures.”
Looking at how AI can help with specific challenges at the checkout, Doorty says that AI tools can determine if an item’s barcode truly belongs to the product being purchased, stopping ticket switching. This way, a shopper’s behavior can be detected and identify suspicious behavior. Once suspicious behavior is detected, an automatic alert can be sent to the SCO attendant.
Doorty added that AI can also detect and identify suspicious behavior related to scan avoidance and other checkout scams, and using UHF RFID as a sensory input for the AI engine to sense the actual product being moved through the checkout area can strengthen your defense even further.
Though cost is of course a valid concern for retailers when it comes to investing in this technology, there are ways to make implementation more cost-effective. Hoang suggests a “try and buy” approach.
“By integrating AI-powered loss prevention into existing infrastructure, retailers won’t need to invest in expensive, fragmented solutions,” Hoang explained. “This approach ensures that protecting profits doesn’t come at the cost of efficiency, scalability, or affordability. AI-powered loss prevention isn’t just a theory; it’s a proven, data-driven approach that can be tested in your own store with minimal risk.”
Other advice our sources offered for retailers regarding AI included:
- Evaluate your brand promise and specific needs. AI checkout solutions are not one-size-fits-all, and different segments and brands will have different needs and goals.
- Focus on ROI. Take an expansive view of the costs at the checkout, including things like square footage and operational costs, peak seasons, etc., to figure out the right mix of staffed, self, fixed, and mobile checkouts.
- Consider customer needs. Make sure your customers are ready for digital experiences and lean into them carefully. It’s a journey.
- Work with your IT department to build the infrastructure and backend when implementing AI in-store.
- Partner with other executives so that they will support you when it comes time for budgeting.
- Understand how you can collect, aggregate, deliver, and respond to the data. AI cameras and sensors track customer interactions, self-checkout behavior, engagement levels, and transaction data, ensuring accurate insights. All this information is typically processed in a centralized platform, consolidating video analytics, transaction monitoring, and AI-driven alerts into a single, easy-to-use system. This can all help ensure employees engage at the right moments, preventing fraudulent scans, and even predicting high-risk situations before they happen.
Looking to the Future
While AI is already making a massive impact on the retail industry, it’s clear that its influence will only continue to grow.
“AI will evolve to not just react to theft at the checkout, but to predict and prevent it before it happens,” Hoang said. “Over time, as more data is collected and consumed by the AI engine on customer behavior, transaction anomalies, engagement levels, and self-checkout patterns, predictive modeling is being applied to the solution. This means retailers will be able to anticipate potential risks before they occur, for example, knowing when and where self-checkout fraud is most likely to happen based on past patterns. AI can then alert staff ahead of time, adjust security measures dynamically, or even automate responses, such as requiring manual approval for high-risk transactions. By integrating real-time AI insights with predictive analytics, retailers won’t just reduce loss—they will be able to fine-tune store operations, improve customer flow, and enhance overall efficiency, making loss prevention both proactive and cost-effective.”
And as criminals become more clever in their fraud techniques, AI will also continue to improve.
“AI products will get more accurate, faster, and be able to solve more and more cases,” Guiste said. “Fraud techniques will grow in sophistication, so the tools to mitigate them will also evolve.”
Langley adds that while store managers are wary of new tools because they might seem like just another thing to manage, AI is well worth the effort. For those hesitant, she recommends starting with a pilot program and implementing AI in a few high-risk stores to assess its impact before scaling.
“AI is not just a tool for loss prevention—it’s a way to future-proof retail operations,” she explains. “As retail technology advances, AI will become a standard part of smart checkout solutions, improving security while enhancing the customer experience. An important consideration is to always look for solutions that will scale as your company grows. Look for solutions that are easy to deploy at a low cost but are high quality. These will be solutions that offer a high degree of security, flexibility for building into existing data or camera systems, low long-term maintenance costs, and strong customer support to ensure smooth deployment.”
