The Organisation for Economic Co-operation and Development has published a paper examining how artificial intelligence and Open Finance interact in financial services. It finds a mutually reinforcing relationship in which Open Finance can provide the richer, interoperable data needed for AI model training, personalisation and broader financial inclusion, while AI can help financial firms and data-sharing systems extract more value from shared data through analytics, automation and standardisation. At the same time, the paper says combining the two trends can amplify existing risks around bias, transparency, data governance, privacy, cybersecurity and consumer overreliance on automated outputs. The paper highlights several practical tensions for policymakers and firms. These include balancing model performance against data minimisation, managing consent in the face of repeated AI model training cycles, and avoiding fragmentation from data localisation requirements. It also discusses trade-offs around the participation of large technology firms in Open Finance frameworks, noting potential gains for innovation alongside concerns about concentration and third-party dependency. In a forward-looking theoretical scenario, the paper says Agentic AI combined with Open Finance data could support more proactive and hyper-personalised financial management, but would also raise sharper questions on accountability, liability, human control and the risk of function creep. As preliminary policy considerations, the paper points to robust governance, human oversight, clear accountability, strong data security and explicit customer consent as core safeguards for responsible deployment. It also calls for closer co-operation across financial, AI, privacy and competition authorities, a proportionate risk-based approach to implementation, stronger consumer awareness, and a possible fact-finding exercise on how existing data-sharing systems are operating in the context of accelerating AI development.

Original source Back to tracker