The article discusses rising AI-enabled fraud threats in retail, highlighting cases like gift card theft and returns fraud, particularly with the Universal Commerce Protocol (UCP). By 2030, agentic AI could manage 15-25% of e-commerce, potentially revamping retail practices. However, security risks are notable, as one in four data breaches by 2028 may stem from AI misuse. The blog outlines specific attack methods, such as hidden instructions in metadata for unauthorized refunds or gift card theft during checkout processes. These techniques pose significant risks to customer loyalty and retailer reputations, resulting in enormous financial losses. Proper safeguards like Know Your Agent (KYA) and through organizations such as the NRF Center for Digital Risk & Innovation are essential in building trust and mitigating risks. Palo Alto Networks offers tools, including Unit 42 AI Security Assessments, to help retailers combat these evolving threats to their digital infrastructures.
Source link
