Retailers are calling on AI to stop customer return fraud

The retail industry is facing an increasing challenge with return fraud, prompting merchants to seek innovative solutions. Common scenarios involve customers ordering multiple sizes of clothing, returning items that don't fit, and even swapping labels or returning worn items. According to Juan Hernandez-Campos, COO of Happy Returns, a company focused on managing returns, the sophistication and scale of return fraud are on the rise. Currently, about 9% of all returns are fraudulent, as highlighted in the National Retail Federation's 2025 Retail Returns Landscape report. This growing issue is forcing retailers to abandon outdated, manual return processes in favor of advanced technologies. AI is emerging as a critical tool in managing reverse logistics, the sector of the supply chain responsible for handling returns and restocking. Companies like Narvar and Loop are leading the charge by employing AI to analyze vast amounts of consumer data, aiming to identify and prevent fraud. Jackie Swanson from Gartner Consulting noted that there has been a significant shift towards AI systems over the last year and a half, which are designed to not only prevent fraud but also expedite legitimate returns, ultimately getting inventory back on the sales floor more quickly. Retailers with annual revenues exceeding $1 billion are increasingly adopting AI solutions, particularly in sectors such as apparel, beauty, and footwear, which experience higher return rates. Caroline Reppert, a senior director at the National Retail Federation, emphasized that reverse logistics is becoming a focal point for retailers exploring AI applications. The types of return fraud include overstated returns, where consumers claim to have returned more items than they actually did, as well as empty-box returns and even counterfeit product substitutions. One sophisticated form of fraud is 'wardrobing,' where customers wear items briefly before returning them. AI's chief advantage lies in its ability to process extensive return data to uncover patterns that might go unnoticed in manual reviews. Happy Returns utilizes a tool called Return Vision, which detects irregularities such as incorrect logos or alterations in product tags. When discrepancies are identified, the system provides real-time alerts to retailers, which helps them address potential issues proactively. Additionally, AI-driven behavioral risk scoring allows retailers to pinpoint high-risk returns based on historical data, enabling them to flag certain returns for human evaluation before issuing refunds. However, despite the advantages, a mixed perspective on AI's effectiveness in combatting return fraud remains, with only 45% of surveyed companies believing that AI alone can solve the problem. For instance, Bully Max, which produces dog food and supplements, employs AI from Shopify and Chargeflow to detect fraudulent returns. CEO Matthew Kinneman acknowledges that while AI assists in generating evidence for disputes, human oversight remains essential in the process. AI's utility extends beyond just fraud prevention; it also helps retailers understand return patterns, allowing them to enhance product descriptions to reduce return rates. Swanson mentioned that her clients are leveraging AI for predictive analytics, identifying orders likely to be returned before the sale is finalized. As returns have traditionally been a cumbersome process, AI technology allows retailers to manage higher volumes more efficiently and consistently. This improved processing not only facilitates faster refunds for customers but also fosters brand loyalty by enhancing the overall return experience.