Finance leaders are finally coming around to AI. They have traditionally been cautious about relying on systems they don’t fully understand – and rightly so. But that is changing as their confidence in more advanced AI systems grows.
In fact, CFO perceptions of generative AI’s importance for financial planning and analysis has nearly doubled this year. Nine in ten finance chiefs now deem it “very or extremely important,” up from 47% in March 2024, according to a new report by Pymnts Intelligence, a data platform. This signals a growing operational dependence on generative AI across core financial functions, such as reporting, cost management and strategic planning.
But AI systems can go far beyond crunching numbers and automating spreadsheets. From deciphering behavioural cues to decoding trader slang, AI’s role in finance is expanding in new and unexpected directions.
Here are five surprising ways it is reshaping the financial world – for better or worse.
Boosting market resilience
Navigating volatile markets and shifting consumer demands is a constant challenge facing finance leaders. But innovative AI tools are emerging to help with consultancy projects. By analysing vast datasets to find patterns in a company’s operations, supply chain or customer base, these systems can surface hidden value in the business and reveal opportunities for diversification – before market pressures force a reaction.
At Nottingham Business School’s Centre for Business and Industry Transformation, researchers have developed a system that analyses the potential value of companies’ competencies, rather than their current products, and suggests how they can be redeployed into entirely new markets.
One example of it in use was to advise a company that made car seats. The tool highlighted their proficiencies in metal processing, allowing the team to pivot into becoming a jewellery exporter. When their market reached saturation, they did the same again, pivoting to make metal wound clamps for battlefield medics.
Translating finance speak
Financial conspiracies are built on jargon-laden, often coded exchanges between crooked traders when they know their telephone conversations and emails are being recorded. But large language models are being developed by compliance providers that can decode clandestine communications in the battle against financial crime.
Banks, hedge funds and regulators can use these tools to track employee messages and uncover secret languages. One company claims its pilot system can decipher phrases coded with emojis and even pig Latin.
Elsewhere, generative AI is finding a different niche in the boardroom, producing natural-language narratives that translate complex financial performance data into clear, accessible summaries. This frees up finance leaders’ and analysts’ time and makes reports more digestible for non-financial stakeholders.
Detecting emotion
The world of finance is not known to be warm and fuzzy. But an emerging field of AI is centered on helping bankers and other finance professionals connect with customers on a more personal level.
There are companies building systems capable of understanding and responding to human emotions. They can predict how a specific word, phrase or pause might trigger feelings of frustration or confusion. As conversations unfold, the user is given prompts, guiding them on how to respond more effectively.
Traditional customer service tactics in finance rely heavily on demographic data, such as age, income and risk tolerance, to suggest products and strategies. Emotional analytics, however, adds a layer of immediacy, enabling banks or finance advisors to adjust their advice based on how a customer is feeling at that very moment.
The application of these systems in finance extend far beyond customer care. In procurement or negotiation calls, for instance, analysing tone and speech patterns could help teams spot risk factors, such as hesitation, tension or evasiveness. Similar tools are also being developed specifically to serve vulnerable people more effectively by flagging vocal cues that might indicate dementia or a mental illness. This could help finance professionals implement relevant safeguards.
The global emotion detection and recognition market is forecast to reach $136bn by 2030, fuelled by innovations in deep learning algorithms.
However, the integration of emotional AI in financial services is not without its challenges. Chief among them are the ethical and privacy concerns that arise when firms begin deploying Orwell-worthy surveillance tech and committing mass emotional manipulation. In the UK, the House of Commons Treasury Committee’s inquiry on AI in financial services recently received written evidence that both lauded emotional and behavioural surveillance and warned about its harmful effects.
Predicting burnout
Burnout looms large in the finance function. Almost all (99%) accountants say they have experienced burnout as a result of unrelenting workload pressures, according to a study last year by accounting software firm FloQast.
Some companies are enlisting AI as an early-warning system, scanning workplace data to flag signs of stress or overwork before it spirals. By analysing expense claim patterns, for example, AI can spot the tell-tale markers of exhaustion, such as frequent late-night meals or a surge in last-minute travel bookings.
Global consumer goods giant, Unilever, has already rolled out an AI-powered wellbeing programme that monitors workplace data for potential burnout risks. It has been integrated with Microsoft 365 suite to analyse calendar and email patterns while maintaining strict privacy controls.
While companies may argue that the technology supports employee wellbeing, in practice it often resembles intrusive workplace surveillance. If not implemented carefully, these systems could end up having the opposite effect on workers’ mental health and morale.
Hunting down fraud
It was just last week that Sam Altman warned that AI is on the edge of an “impending fraud crisis” in which “anyone will be able to perfectly imitate” anyone else. It is something that finance leaders are acutely concerned about. Last year, a deepfake CFO scammed the British design firm Arup out of $25m (£19m) – a chilling reminder of the increasingly sophisticated attempts to defraud companies using phishing scams and false videos.
To combat this, some firms are trialling AI systems that go beyond flagging suspicious transactions and can analyse the behavioural biometrics of users.
During high-risk transactions, for example, these systems can detect if a user is typing erratically, hesitating, switching devices or appears to be coached over the phone. Such patterns can indicate social engineering or coercion in real time.
Some tools even have a daily adaptive model that retrains itself every 24 hours based on new behavioural data and fraud signals. This allows it to detect emerging threats without human reprogramming.
Finance leaders are finally coming around to AI. They have traditionally been cautious about relying on systems they don’t fully understand – and rightly so. But that is changing as their confidence in more advanced AI systems grows.
In fact, CFO perceptions of generative AI’s importance for financial planning and analysis has nearly doubled this year. Nine in ten finance chiefs now deem it “very or extremely important,” up from 47% in March 2024, according to a new report by Pymnts Intelligence, a data platform. This signals a growing operational dependence on generative AI across core financial functions, such as reporting, cost management and strategic planning.
But AI systems can go far beyond crunching numbers and automating spreadsheets. From deciphering behavioural cues to decoding trader slang, AI’s role in finance is expanding in new and unexpected directions.
