UN digital tech agency launches initiative to improve trust in AI agents (www.reuters.com)

The United Nations' International Telecommunication Union (ITU) has launched a new initiative aimed at enhancing trust in increasingly autonomous AI agents, which are designed to operate independently in tasks like scheduling and complex business activities. This initiative comes in response to growing concerns over the accountability and oversight of these systems, particularly their potential to impersonate users and make unauthorized decisions. To tackle these risks, the ITU announced the formation of a Focus Group during the AI for Good Summit in Geneva. This group will consist of experts in technical, policy, and legal fields, and its primary goal is to develop frameworks that ensure AI agents are identifiable, trustworthy, and under meaningful human control, especially in sensitive areas such as finance and critical infrastructure. Co-Chair Debora Comparin emphasized the need for international standards to determine when AI agents can be trusted to negotiate and make decisions on behalf of users. The group will convene for its inaugural meeting in Paris this November.