The New Retail Arms Race: AI vs. AI in the Battle Over Returns

NEW YORK, NY – June 22, 2026 – A quiet but significant battle is escalating across the digital aisles of e-commerce, and the weapon of choice is artificial intelligence. A structural shift in consumer behavior, detailed in a new global report by the e-commerce risk intelligence firm Riskified, reveals that nearly half of all consumers are now using generative AI tools like ChatGPT to draft and dispute return claims. This democratization of sophisticated persuasion is fueling a surge in policy abuse, forcing retailers into a defensive, technology-driven posture to protect their razor-thin margins. The investment landscape, already grappling with the complexities of AI integration, must now account for its role as both a tool for fraud and a shield against it.

The Consumer's New Toolkit: Normalizing Abuse

The report, “Rewriting the Rules on Returns,” paints a stark picture of a consumer base increasingly comfortable with bending, if not breaking, the rules. It found that a staggering 85% of shoppers now find at least one form of strategic or borderline return behavior acceptable. These aren't just minor infractions. According to the research, 46% of consumers think it's perfectly fine to return an item simply because it looks different in person than it did online. Another 42% admit to “bracketing”—the practice of buying multiple sizes or colors with the full intention of returning most of them. Perhaps most alarmingly for retailers, 24% believe it is acceptable to wear or use an item before sending it back for a full refund.

This trend is being amplified and normalized by a social media culture awash with “life hacks” that provide informal guidance on how to exploit generous return policies. More than half of consumers report encountering such content online. The introduction of generative AI into this equation acts as a powerful accelerant. Shoppers are no longer just sending a quick email; they are using AI to generate highly convincing, well-written, and often difficult-to-refute return requests. As one retail leader interviewed for the study noted, these AI-crafted claims are overwhelming manual review teams, making it harder to distinguish legitimate issues from calculated abuse.

A High-Stakes Financial Drain

While consumers may view these tactics as levelling the playing field, the collective financial impact on the retail sector is immense. The cost of managing returns, refunds, and exchanges is no longer a simple cost of doing business; it is a structural threat to profitability. Independent analysis estimates the global cost of returns—factoring in foregone revenue, lost stock, and logistics—at a staggering $394 billion in 2024. The Riskified report adds a crucial layer to this, finding that nearly one in every four dollars claimed in refunds is abusive.

This abuse is not evenly distributed. The data shows a clear pattern of abusers targeting higher-value orders, where the financial incentive is greatest. Claim rates for orders over $2,000 are 2.5 times higher than for those under $100, and orders exceeding $1,000 are 33% more likely to involve an abusive claim. This puts luxury and high-end retailers in a particularly vulnerable position. The operational strain is also significant, with the post-holiday season rush in January seeing a flood of claims from November and December purchases, stretching fulfillment and customer service teams to their limits.

Retail's Counteroffensive: The AI Arms Race

Faced with this onslaught, retailers are abandoning passive acceptance and are now engaged in a full-blown AI arms race of their own. The response is twofold: tightening policies and deploying sophisticated technology. Over the past year, roughly a third of retailers have introduced return fees, while others have shortened return windows to as little as seven days and shifted toward store credit or exchange-only policies. However, these blunt-force measures risk alienating loyal, high-value customers.

This is where technology becomes the critical differentiator. Companies like Riskified, along with competitors such as Forter and Signifyd, are providing the arsenal for this new battlefront. These platforms use AI and identity-based intelligence to analyze the individual behind each interaction, creating a real-time risk profile. “When half of consumers are using AI to draft highly persuasive refund claims, manual reviews simply cannot keep up,” said Jeff Otto, Chief Marketing Officer at Riskified. “To protect margins without alienating top shoppers, retailers need the ability to provide differentiated return experiences, accurately and at scale.”

The results from this approach can be dramatic. A global luxury fashion brand using Riskified’s platform reported a 75% reduction in chargebacks while simultaneously increasing its conversion rate from approximately 50% to over 75%. By accurately identifying repeat offenders across multiple identities and devices, the brand was able to apply stricter controls to abusers while maintaining a frictionless, premium experience for its legitimate customer base.

The Emerging Landscape of Trust and Transparency

The future of e-commerce returns will not be defined by a victory for one side, but by a new, dynamic equilibrium. This evolving landscape raises complex ethical and legal questions. As retailers use AI to create tiered customer experiences, they must navigate a tightrope of data privacy regulations like GDPR and CCPA, while also ensuring their algorithms are free from biases that could be deemed discriminatory.

Surprisingly, consumers seem ready for this new paradigm. The Riskified study found that a majority (52%) are supportive of stricter return policies, and even more (56%) prefer personalized or tiered policies over a one-size-fits-all approach. Furthermore, over 60% of shoppers indicated they would adjust their behavior if they had a better understanding of the true costs associated with returns. This suggests a path forward built on transparency and dynamic, risk-based personalization. The retailers that thrive will be those that can master this delicate balance, using AI not as a wall, but as a sophisticated lens to distinguish their most valuable customers from those who seek to exploit the system.