Six major international financial institutions, including Bank of America and Capital One, have released a joint report titled Building Trust in Agentic Commerce. The document highlights that while AI-driven shopping agents possess significant potential, they introduce critical challenges regarding transparency, security, data privacy, and fraud risks.
The report identifies five core principles for the industry: transparency, security, privacy and data protection, user choice, and platform interoperability. Financial institutions emphasize that the current speed of technological development is outpacing the establishment of regulatory frameworks and consumer protection mechanisms.
Concerns raised by the banking sector include the potential for AI agents to prioritize products with higher commissions or lower transaction costs over the user's best interests. Furthermore, there are risks of unauthorized account access, budget overruns, and potential financial losses due to system vulnerabilities or data breaches.
Recent industry incidents underscore these security concerns. Meta recently patched a zero-day vulnerability in its Muse AI shopping assistant that could have allowed attackers to take control of the agent. Additionally, Amazon has publicly challenged Meta, alleging that its AI agents accessed the platform without proper identification, raising concerns over data handling.
Despite these hurdles, the banking coalition views agentic commerce as a potential future standard for transactions. The group plans to continue research to define industry standards and collaborate with regulators and partners to ensure the healthy development of this technology.
