Retail
AI shopping agents raise new security risks for retailers ahead of holiday season
Thales warned that AI shopping agents are increasing retail risks involving authorization, sensitive-data disclosure, fraud, and accountability. It recommended identity checks, strict permissions, and human confirmation for high-risk actions.
Thales data highlighted security and governance risks associated with AI shopping agents ahead of the holiday season. Retail and e-commerce leaders reported that AI development is moving faster than security teams can manage, creating concerns around sensitive data, authorization, fraud, and accountability.
The report said 72% of retail and e-commerce leaders believed AI was changing too quickly for security teams to keep pace. It also found that 61% reported AI-driven disclosure of sensitive data, while only 37% had a complete view of where their data was located.
The risks increase as agents receive authority to search, recommend, exchange information, and complete purchases. Retailers must determine what actions an agent is allowed to take, how customers authenticate those actions, and who is responsible when an automated decision causes harm.
Thales recommended stronger identity verification, granular permissions that agents cannot override, and confirmation steps for high-risk actions requiring human approval. The issue is particularly relevant during peak shopping periods, when higher transaction volumes can magnify fraud, privacy, and operational failures.
Original source: Retail Insider