AI Reshapes Corporate Treasury: Citi's Vision for Liquidity

Corporate treasuries are on the cusp of a profound transformation, driven by the accelerating integration of artificial intelligence (AI). This new era sees AI becoming an indispensable tool for daily decision-making, optimizing cash positions, and revolutionizing liquidity management. A groundbreaking report by Citi illuminates that while 82% of treasury teams are currently in the nascent stages of experimenting with generative AI, only a mere 3% have successfully scaled its adoption across their operational frameworks. Despite the initial cautious approach, the study boldly projects that by 2030, AI will evolve into "the new treasury operating system," fundamentally altering the function from a traditional control center into a dynamic, intelligent financial hub.

Unlocking Productivity: The Generative AI Advantage

"The potential productivity gains from AI are too significant to ignore," asserts Ron Chakravarti, Citi’s Global Head of Client Advisory Group, emphasizing generative AI as the "getting-things-done tool for treasury." The Citi report delineates a comprehensive four-stage maturity model guiding this evolution. This model commences with the diligent identification of pertinent use cases, advancing through meticulous exploration and strategic transformation, ultimately culminating in continuous optimization. Each stage inherently demands quantifiable results, robustly structured data ecosystems, and unwavering human oversight to ensure effective and reliable integration.

Global Perspectives: Citi’s Survey Illuminates AI Trajectory

Citi’s findings are meticulously grounded in a global survey encompassing 75 corporate treasuries spanning diverse industries and geographical regions. While a majority of respondents remain entrenched in pilot initiatives, the overarching direction towards AI integration is unequivocally clear. Nearly 60% of treasurers report having identified at least one practical use case for generative AI, with a significant 40% planning to escalate their investments in this technology within the next two years. The most prevalent applications currently under exploration include advanced liquidity forecasting, streamlined reconciliation processes, and automated report generation. A smaller, yet pioneering, cohort is actively testing generative AI for sophisticated variance analysis and the creation of narrative components in critical management reports.

Navigating the Data Landscape: A Critical Barrier

Despite the palpable enthusiasm, data quality emerges as the single most formidable barrier to widespread AI adoption. Over 70% of respondents explicitly cited fragmented or incomplete data as a significant constraint. To circumvent this impediment, the report strongly advocates for the establishment of a centralized data lake, forging robust API connections to enterprise resource planning (ERP) systems, and unequivocally defining ownership for data accuracy. Without these foundational pillars, Citi cautions that "AI will only replicate human errors at greater speed," undermining its transformative potential.

"Treasury is the ultimate guardian," states Joseph Neu, Founder and CEO of NeuGroup, adding that "There must be 100% trust in the numbers. Generative AI has been slow to deliver at this level of trust." Citi attributes this prevailing skepticism to a broader cultural challenge within organizations. While AI undeniably enhances speed and precision, its inherent credibility, the report argues, is intrinsically linked to robust governance frameworks, comprehensive explainability, and transparent audit trails.

Fostering Human Readiness for AI Integration

The imperative for transparency resonates deeply with treasurers themselves. Alexander Reijrink, Global Head of Corporate Finance and Risk Management at Philips, underscores their proactive approach: "As a first step, we invested time to train the treasury team and trigger a change mindset. This helps us in finding the most valuable use cases, wherever they come up." Citi frames such initiatives as absolutely essential for cultivating the human readiness and adaptability required for successful, sustainable AI adoption.

Building Towards the AI-Driven Treasury Operating System

The ambitious transformation envisioned by Citi is already manifesting across various segments of the financial ecosystem. Treasury teams are progressively abandoning antiquated manual spreadsheets, pivoting towards sophisticated platforms powered by predictive analytics and advanced data intelligence. Bank of America’s CashPro, for instance, exemplifies this shift by providing treasurers with real-time visibility into global cash positions and predictive forecasts, demonstrating how meticulously structured data empowers faster, more confident financial decisions.

Exploring Agentic AI and Enhanced Automation

Concurrently, experiments with agentic AI are actively pushing the boundaries of automation within treasury functions. Some pioneering firms are developing systems capable of recommending or even executing internal transfers, all while rigorously maintaining human review processes and comprehensive traceability. These innovative prototypes closely align with what Citi describes as the "transformation" stage of AI maturity, where intelligent models augment human capabilities without yet operating fully autonomously.

Citi’s own Treasury and Trade Solutions (TTS) group is actively translating these theoretical concepts into tangible real-world transactions. In its most recent quarterly update, the bank revealed its efforts to extend tokenization and programmable money capabilities to corporate clients. This innovation facilitates instant cross-border liquidity and more highly automated cash management, perfectly aligning with the report’s architectural vision for future treasuries: interconnected through APIs, governed by stringent data standards, and engineered for continuous, real-time operation.

Elevating the Strategic Role of Treasury in the Digital Age

The shift extends beyond mere technical enhancements; it fundamentally redefines the strategic importance of the treasury function. As AI and cyber risk converge, the role of treasurers is significantly expanding. They are evolving into pivotal strategic participants in enterprise planning, overseeing not only traditional liquidity management but also critical areas such as payments infrastructure, comprehensive data quality assurance, and digital resilience. Citi’s findings strongly corroborate this evolution, observing that treasurers who proactively collaborate with technology and data teams from the outset are markedly better positioned to transition from initial experimentation to full-scale transformation.

Prudent Phased Deployment and Continuous Validation

While the trajectory towards AI integration is overwhelmingly positive, a degree of caution remains imperative. Citi unequivocally stresses that the full deployment of AI in treasury should proceed through carefully managed phases, anchored by consistent human validation and the demonstration of measurable outcomes. The report highlights that 61% of surveyed treasurers express a clear preference for initiating small pilot projects that yield quick, demonstrable wins before embarking on broader scaling efforts. The authors issue a crucial warning: premature automation devoid of adequate oversight could paradoxically erode credibility rather than enhance it, underscoring the critical need for a balanced and strategic approach to AI adoption.