The Future of AI in Treasury

TMANY Members Share AI Aspirations, Challenges, and Next Steps in Treasury

The Treasury Management Association of New York (TMANY) recently hosted an engaging luncheon roundtable to explore how treasury professionals are thinking about—and beginning to implement—AI in treasury operations. The event was led by Bridget Meyer, Senior Director at Redbridge Debt & Treasury Advisory, and Daniel Kalish, CEO of Nilus.

What Treasury Teams Want AI to Solve

The roundtable opened with participants identifying high-value use cases they hope AI in treasury can address. Unsurprisingly, improving cash forecasting topped the list, with members eager to incorporate leading market indicators and contrarian insights to strengthen financial planning. Others expressed a desire for one-page balance summaries to support executive visibility and better decision-making.

AI is also being looked to for customer service automation, faster integration of new finance systems, and real-time scenario analysis—especially in response to price or market shocks. Many envision AI as a virtual knowledge resource that can support internal teams and streamline day-to-day operations, including CRM updates, payment processing, and invoice generation.

Additional opportunities for AI in treasury include reconciling contract and invoice terms, aggregating data from disparate platforms, and managing communication overload—particularly through summarizing and prioritizing hundreds of daily emails.

What’s Already Being Tried With AI in Treasury

Several attendees shared how they’ve already begun experimenting with treasury AI tools in their organizations. These ranged from embedding Copilot into Microsoft Word to standardize SOP formatting, to using AI for vendor research, data cleansing, and generating automated trend reports.

Other current applications of AI in treasury include pulling key terms from loan agreements, building presentations for senior management, drafting performance reviews, and translating or summarizing dense webinar and reporting content. Teams are also using AI to help develop relationship scorecards for banks, refine client service strategies, and enhance internal communications by rewriting emails and converting notes into digestible summaries.

Barriers to Adoption

Despite the momentum, attendees noted several common obstacles to broader AI adoption in treasury. Chief among them were budget constraints, unclear ROI, and internal resistance to change—particularly fears around losing the human element or diminishing the value of critical thinking.

Data readiness also emerged as a significant hurdle, as unstructured or inconsistent data can limit the accuracy and effectiveness of AI tools. Participants raised valid concerns about data security, confidentiality, and regulatory compliance—especially when dealing with sensitive financial data.

On a tactical level, treasury teams are also navigating a crowded AI vendor landscape, limited time to evaluate tools, and a general lack of education around prompt engineering and implementation strategy. Larger enterprises often restrict AI experimentation, while smaller firms provide more flexibility—creating a divide between early adopters and those lagging behind.

Next Steps and the Future of AI in Treasury

Participants agreed that developing a formal AI policy is a top priority to define internal guardrails and ensure responsible use. Others committed to changing daily habits—like replacing Google with AI assistants—and promoting a growth mindset: AI is not a threat, but a powerful tool to be embraced.

Immediate next steps also included upskilling through prompt engineering courses, leaning on internal AI champions, and aligning with peers and vendors who are actively embracing AI in treasury. Some also plan to review third-party AI practices to ensure they align with internal standards.

As treasury teams continue exploring this fast-evolving space, one message rang clear: start small, stay informed, and don’t wait to engage.

A Note from Your Co-Author

As you’ve explored the cutting edge of AI in treasury, you might appreciate a little meta-insight: this article itself was generated by AI. Using iPhone Voice Recorder transcripts as a foundation and ChatGPT for refining the language, an AI assistant brought these insights to life with minimal human intervention. And yes, even the article’s thumbnail was created by AI. It’s a testament to the very tools we’re discussing—showing how AI can truly augment human efforts, making complex topics more accessible and engaging, right down to the visual elements. Just another powerful example of AI at work, helping us share valuable information.