Citi: Treasuries Are Building Momentum In AI Adoption And Liquidity Transformation

Core Insights - Corporate treasuries are entering a new era where artificial intelligence (AI) is becoming integral to daily operations, cash forecasting, and liquidity optimization [1][3] - A Citi report indicates that 82% of treasury teams are in the early stages of experimenting with generative AI, with only 3% having scaled its adoption [1][4] - By 2030, AI is expected to evolve into "the new treasury operating system," transforming treasury functions into intelligent financial hubs [1] AI Adoption and Maturity Model - The report outlines a four-stage maturity model for AI adoption in treasury, starting from identifying use cases to exploration, transformation, and optimization [3] - Nearly 60% of treasurers have identified at least one practical generative AI use case, and 40% plan to increase investment in AI within the next two years [4] Data Quality and Challenges - Data quality is cited as the biggest barrier to AI adoption, with over 70% of respondents indicating fragmented or incomplete data as a major constraint [5] - Recommendations include building a centralized data lake and establishing API connections to improve data accuracy [5] Human Readiness and Training - Companies are investing in training their treasury teams to foster a change mindset, which is essential for identifying valuable AI use cases [6] - Human readiness is framed as crucial for successful AI adoption [6] Technological Transformation - Treasury teams are moving away from manual spreadsheets to platforms powered by predictive analytics and data intelligence [7] - Examples include Bank of America's CashPro, which provides real-time visibility into global cash positions [7] Strategic Role of Treasury - The role of treasury is expanding to include oversight of payments infrastructure, data quality, and digital resilience, as AI and cyber risk converge [11] - Treasurers collaborating with technology and data teams early are better positioned for transformation [11] Caution in Implementation - Full AI deployment in treasury should proceed in phases, anchored by human validation and measurable outcomes [12] - 61% of surveyed treasurers prefer starting with small pilots to demonstrate quick wins before scaling [12]