Retailers have significantly tightened oversight of employee AI tool adoption, with the share of retail workers using ungoverned AI tools declining from 70% to 44% since Netskope's previous annual report. Meanwhile, adoption of officially approved tools surged from 40% to 73%, and overall two-thirds of retail employees now use some form of standalone AI tool.
Despite these gains in shadow AI reduction, retailers face mounting challenges in controlling where sensitive data flows. Nearly all retail employees use software containing AI features, and 90% reported using AI tools trained on customer data. Of AI-related data policy violations tracked by Netskope, 56% involved regulated customer data. Agentic AI poses a particular visibility blind spot: Netskope observed a 400% increase in retail AI agents interacting with remote Model Context Protocol (MCP) servers during the survey period, creating additional pathways for unauthorized data movement.
Retailers can mitigate these risks by blocking unnecessary applications, inspecting web traffic, and deploying data-loss-prevention policies to detect sensitive information flowing to ungoverned tools. The report, based on data collected between July 2025 and July 2026, highlights the tension between rapid AI adoption and governance maturity in the sector.