Leveraging generative AI in banking treasury functions

The applications for generative artificial intelligence (gen AI) are gaining traction across all industries, and financial services are no exception. From fraud prevention and chatbots to analyzing environmental, social and governance (ESG) data, the technology’s use cases are vast and reshaping the industry as we know it. That includes potential within banking treasury management.

Technology and automation are crucial in a modern treasury function, enabling banks to replace their manual processes with sophisticated software solutions, applications and automation tools to streamline operations, improve efficiency and ultimately enhance decision-making processes. With its ability to learn from data and produce new content, gen AI is today becoming a big driver of efficiency and automation, creating new opportunities for treasurers to better analyze risk, evolve with change, make more informed decisions and generate new opportunities for growth.

Applications of gen AI in banking treasury

One major use case for gen AI in a bank’s treasury department is the ability to analyze, predict and navigate market trends. Financial markets can move extremely quickly, and the window to take advantage of opportunities while mitigating risks can be small. Particularly as black swan events become more frequent, treasurers are navigating more hurdles than ever before from pension performance, interest rates and inflation, global pandemics and major political unrest.

For example, when Covid-19 took place, banks needed to quickly understand sentiment and positions in the market, assess potential impacts on liquidity and investments, and make fast decisions around hedging and more. Banks need to navigate disruptions and take advantage of growth opportunities within their treasury operations and this is where AI, and gen AI in particular, play a big role.

By ingesting large volumes of historical data and analyzing current events that may impact decision-making – including decisions by the Federal Reserve, trending news or social media -driven sentiment – gen AI can help to predict market movements in real-time, and with an increasing amount of accuracy. Treasurers can then make more informed decisions, quickly, about liquidity and cash management requirements, how to adapt investment strategies and mitigate risk. Gen AI also provides better trends analysis about a bank’s customers, their behaviors and cash flow positions.

Fraud detection is yet another strong use case. Fighting fraud remains a top priority for banks particularly as services and functions increasingly operate in real time, leading to increased risk of criminal behavior. AI and gen AI potentially strengthen a bank’s ability to identify suspicious transactions and fraudulent activity in real time as well. Through synthetic data, banks can analyze and plan for future fraud scenarios that may not have happened before. Ultimately, it means financial institutions minimize losses and ensure greater protection for their assets as well as their customers.

The technology is not without challenges

When it comes to using gen AI for real-time data analysis and decision-making, there are some challenges. For one, gen AI is only as good as the quality of data feeding it. Treasury traders or managers demand confidence in gen AI to make timely and high-stakes decisions. Banks also need teams of people – both externally and internally – who can understand the decisions made by AI, how it works and the patterns it identifies, and then test the application on an ongoing basis to ensure continued accuracy.

There are other challenges around data governance, regulations and privacy, particularly when we consider the data flow throughout a bank’s treasury management ecosystem. This starts with treasury, flows through to middle-office risk management and then back-office processing, with IT at the center of it all. In a highly regulated market, ensuring robust security and data privacy throughout all processes and departments is critical. This is why banks must work with technology providers that can help them do this safely, in a way that is compliant with both internal policies and external regulations.

Finally, we must remember that the market for AI is still maturing, and regulators are still finding their feet. As banks and credit unions implement gen AI as a strategic direction for the treasury function, open conversations with regulators and compliance teams are crucial to ensure they do this in a safe way for the institution and its clients. The rollout of any new technology needs to be embedded in strong governance and gen AI can’t be adopted as a “big bang” approach. As the market develops, new systems, software and regulations will continue to develop too.

Collaboration is crucial for success

Whether it’s cash flow forecasting, automating routine tasks, predicting future trends, detecting anomalies or fraud detection, the applications for gen AI are diverse and transformative, enabling treasury departments to boost efficiency and focus on more complex tasks that require human expertise.

In the coming years, we can expect to see a lot more innovation and sophistication in AI capabilities, with solutions addressing even more complex use cases and unlocking new opportunities in treasury and overall banking operations. To overcome ongoing challenges as the industry evolves, banks, technology companies and regulators must work together to bridge the gaps between serving customers and meeting regulatory requirements.

Collaboration is also key to ensure successful use of gen AI in treasury. Banks must find the right partners that bring tech solutions that are robust, secure and drive genuine value. Banks must avoid adopting the latest technology simply for technology’s sake.

Meggie Grimaud is Head of Analytics, Treasury & Capital Markets at Finastra.