Who's really behind the login? AI agents are forcing a rethink (www.techradar.com)

The increasing use of AI agents in business transactions is prompting a critical reassessment of digital trust, which has traditionally relied on the assumption that human identity is behind every login or action. Current authentication methods, such as passwords and multi-factor authentication, effectively confirm user identity but fall short when it comes to validating the authority under which AI agents operate. This gap highlights the pressing need for businesses to rethink their security infrastructures, as these agents manage tasks without a clear understanding of their authorized scope, potentially leaving organizations vulnerable to unauthorized actions. To address this issue, experts recommend businesses adopt a proactive approach to verifying the authority of AI agents before they initiate actions. A robust framework should be established to clarify who, or what, is acting, who authorized the action, the scope of permissions granted, and whether such authority remains valid. This framework could leverage existing models, such as the ‘on-behalf-of’ delegation, to create machine-verifiable mandates for agents. With regulations like the Data (Use and Access) Act paving the way for enhanced digital identity standards, there is a unique opportunity for the UK to strengthen its digital infrastructure to ensure AI agents operate with clearly defined and limited authority.