In remarks at the AI & DATA Summit, Portuguese Insurance and Pension Funds Supervisory Authority President Gabriel Bernardino called for governance and supervisory frameworks to keep pace as artificial intelligence moves from supporting decisions to autonomously executing complex financial processes. He warned that consumer facing agents could effectively perform financial intermediation while their developers and platform providers remain outside the traditional supervisory perimeter, creating uncertainty over responsibility, incentives and consumer protection. Bernardino said financial institutions and their boards must remain accountable for AI outcomes as system autonomy increases. This requires reliable data, transparency, nondiscrimination, continuous testing and meaningful human oversight designed to prevent approval from becoming a formality. He also highlighted systemic risks from reliance on a small number of technology providers and frontier models, as well as the potential for similar agents to make synchronized decisions at high speed. Against this backdrop, the authority plans to combine controlled experimentation through a future Technological Free Zone with greater use of AI, advanced data analysis and supervisory technology under its INOVAsup program, building on its responsible innovation initiative and strong market support for a testing environment.
Portuguese Insurance and Pension Funds Supervisory Authority president calls for stronger accountability and oversight of autonomous AI agents
At the AI & DATA Summit, Portuguese Insurance and Pension Funds Supervisory Authority President Gabriel Bernardino called for stronger accountability, data governance and meaningful human oversight as autonomous AI agents enter financial services. He warned that such agents could create supervisory perimeter and systemic risks, while outlining controlled testing and greater use of supervisory technology as elements of the authority’s response.