Tomorrow’s Treasury Today: The Pitfalls of AI in Treasury

Having discussed the use of AI and how it has significantly transformed treasury functions by streamlining processes, enhancing efficiency, compiling financial data, and offering deep analytical insights, several areas still require further development and present potential workflow challenges.

Technologies like ChatGPT underscore the potential for job displacement in sectors like software engineering, coding, trading, and journalism, as evidenced by Google’s recent tests that show the software’s capability to pass a job interview.

As this technology continues to evolve and reshape our daily work routines, a critical question arises: what will be the impact of AI’s rise on the treasury sector?

The Association of Corporate Treasurers (ACT) has emphasised that the pace of development in work-related technologies is often slowed by the stringent protocols surrounding cybersecurity. As these technological platforms expand, the associated risks also increase.

The World Economic Forum investigated the threats AI could pose, surveying Chief Risk Officers (CRO) to find out more.

Three in four security professionals say the top three factors in increased cyber risk are AI, geopolitics, and remote working with the finance industry in the top 3 departments most impacted. 

It’s essential to note that the use of AI to target sensitive data is not inherently malignant. AI can also be employed for legitimate purposes, such as enhancing cybersecurity defences, encrypting data, or detecting anomalies in system behaviour to prevent data breaches. 

However, the potential for misuse and malicious intent underscores the need for robust security measures, regular vulnerability assessments, and employee education on data protection and cybersecurity best practices. Organisations should implement a multi-layered security approach to safeguard sensitive data from AI-driven threats.

What to be aware of when using AI in treasury;

• Cyber attacks – As the threat of cybercrime increases with the development of new technology, AI systems, especially those with machine learning capabilities, can analyse vast amounts of data, extract patterns and trends in human behaviour and identify vulnerabilities in computer systems, applications, and networks. Treasury departments are particularly vulnerable to cyber attacks as the core function of treasury is cash management and treasurers “hold the keys to crucial bank accounts” making them the prime target for cybercriminals.
• Data protection – Data security is the foremost challenge AI must tackle in the next decade. Several AI models store information and data that it’s given in order to learn and develop its capabilities. Many treasury departments handle confidential information such as mergers and acquisitions, strategic financial planning and sensitive negotiations. One of the biggest obstacles when implementing AI technology in security is the potential for false positives. Although alternative AI models with internal firewalls exist, this is a fundamental area of artificial intelligence that requires much development.
• Prejudice – Bias in algorithmic decision-making refers to the presence of unfair or prejudiced outcomes caused by the algorithms used in automated systems. These algorithms are designed to process data and make decisions or predictions based on patterns and correlations in the data they have been trained on. However, if the data used to train these algorithms contain biases or reflect historical discrimination, this can affect risk assessments and financial outcomes as the algorithms learn and perpetuate biases in the decision-making process.
• Regulation – As artificial intelligence continues to develop exponentially the regulations and security protocols surrounding it are still in their infancy. Countries worldwide have started creating their own regulations and frameworks to address the legal, ethical, and societal implications of AI.

What’s the future for Treasury Tech and AI? 

AI has firmly established itself in modern society and business operations, in the same way the internet did in the 1990s, permeating all sectors, including treasury management. The growing role of AI in business finance only emphasises the need for enhanced computer literacy in treasury professionals. Advancing technology will bring more sophisticated artificial intelligence and machine learning, further revolutionising the field. 

The establishment of software such as blockchain (distributed ledger technology) is facilitating increasingly secure methods of cash movement and helping to reduce payment fraud, identify theft, protect sensitive data and provide digital transparency and traceability. 

As AI continues to offer unparalleled benefits in terms of efficiency, forecasting, and strategic decision-making, treasurers must adopt it responsibly and ethically. Vigilance about data privacy, security, biases and potential inaccuracies is paramount.

Our Head of Treasury recruitment, Guy Middleton, discusses the impact AI will have on the treasury function and its hiring practices, saying:

“Advancing technologies have been changing the treasury workplace for a long time, and with Artificial Intelligence increasingly being used within treasury systems, it is a given that AI will have a significant impact on the skill sets required within future treasury functions. Treasurers frequently need to question the data they are presented with, meaning that strong mathematical abilities and an analytical mindset are likely to remain essential; although as more tasks become automated, the role of the treasury professional is likely to transition into more of an advisory role involving increased business interaction and strategic input. As a result, the ability to accurately interpret data and clearly present relevant information will be key.” 

While artificial intelligence demonstrates remarkable proficiency in trading and predictive analysis, its position within the treasury market is not one of replacement but rather of collaboration with human expertise. By thoughtfully integrating AI into their operations, treasury professionals can unlock its intrinsic advantages to optimise workflows, streamline task efficiency, and propel methodologies forward. However, human oversight remains essential to validate AI’s outputs, safeguard data integrity, and mitigate biases.