The financial services industry has “no choice” but to adopt and invest in AI-powered cybersecurity solutions, says Kevin Levitt, global director of financial services at Nvidia.
The comments come as the chipmaker releases a new survey which finds that 98 per cent of financial services management will further increase AI spending in 2025.
Based on a survey of approximately 600 global financial services professionals about the trends, challenges, and opportunities for accelerated computing and AI, Nvidia’s fifth annual state of AI in financial services report shows cybersecurity experienced the highest YoY growth within the financial services industry, with more than one-third of respondents now assessing or investing in AI for cybersecurity.
Financial services institutions are particularly driven to invest in AI-powered fraud detection and cybersecurity tools because the technology has proven to be highly effective at reducing the massive problem of transaction fraud, Levitt says.
“AI is a crucial and non-negotiable requirement for effective cybersecurity in the financial services industry going forward, with the move being necessary to keep pace with and outmanoeuvre the AI-enabled tactics being used by malicious actors targeting the financial system,” Levitt told FStech ahead of Nvidia’s report launch.
Because of this, failing to adopt AI for cybersecurity would leave financial firms vulnerable to these advanced threats, he added.
According to the study, the number of respondents expecting to use AI to address spear phishing attacks has more than doubled in the past year, jumping from seven per cent to 17 per cent, signalling a shift in the cyberthreat landscape.
Similarly, the use of AI to confront supply chain attacks and Distributed Denial of Service (DDoS) incidents increased, indicating a heightened awareness and proactive stance against these threats.
“Fraud is a massive $43 billion global problem, particularly in transaction fraud,” Levitt explained, adding that banks, credit card issuers, and payment networks leveraging AI to fight transaction fraud incur in false positives’ reduction while increasing the accuracy of fraud detection.
“It is crucial that financial firms continue to invest in leveraging AI and advanced techniques like deep learning and generative AI to identify fraud, with the focus being on anomaly detection, identifying rare instances that can have meaningful impacts, as new fraud techniques emerge,” continued the global director.
According to Levitt, merchants are also benefitting from AI-powered fraud detection reducing fraudulent transactions, as it leads to more proven, legitimate transactions.
Additionally, within anti-money laundering (AML) and know-your-customer (KYC) processes, financial institutions are using AI to verify customer identities, with genAI capabilities enabling firms to assist in generating critical reports more efficiently and with greater accuracy.
“This is why we are seeing such significant growth in the use of AI and genAI for a wide range of cybersecurity challenges, including defending against ransomware and malware, protecting against spear phishing attacks, and securing against credential and identity-based attacks,” he said.
GenAI at the forefront
According to the report, 41 per cent of management-level respondents now recognise AI and genAI as transformational forces within their organisations. Financial institutions are integrating AI across business functions, with significant upticks in AI use for risk and compliance, marketing, sales, cybersecurity, and operations.
Nvidia’s experts say that techniques like domain-adaptive pretraining, fine-tuning, and retrieval-augmented generation (RAG) are now being used in combination with open-source foundation models to create a flywheel for generative AI development, which increases accuracy while protecting enterprise information.
Half of management respondents indicated that their first genAI service or application had already been deployed, with an additional 28 percent planning deployment within the next six months.
The top genAI use cases in terms of return on investment (ROI) are trading and portfolio optimisation, which account for 25 per cent of responses, followed by customer experience and engagement at 21 per cent.
Agentic AI
Levitt highlighted the growing adoption of AI agents deeply integrated into financial services operations, as well as the increasing investment in AI factories to enable the scaling of AI applications across the industry.
He expects the next development in genAI will be agentic AI, in which financial institutions will use sophisticated, autonomous AI agents for tasks such as cybersecurity threat detection, customer service, and accelerated investment analysis.
“AI agents will continue to be used to improve customer experiences by providing support to call centre agents or engaging customers directly,” said Levitt. “They will also help with compliance and risk mitigation by being trained on current and new regulations to assist banks in meeting regulatory standards.”
According to the report, companies are also taking proactive steps to build AI factories, especially those developed to accelerate computing platforms equipped with full-stack AI software.
“In order to retain this AI talent, banks need to provide the right tools and infrastructure, added Levitt. “This includes investing more in accelerated computing platforms and AI factories, providing more training to their teams to move from traditional machine learning to deep learning and generative AI modelling capabilities.”
This includes the hardware, platform software, and application frameworks to enable data scientists to build and deploy AI applications quickly and cost-effectively, he continued.
“Leading banks are moving from hundreds of AI use cases to thousands, impacting every function and line of business. To build these AI-enabled applications at scale, productively and reliably, banks will invest in AI factories,” Levitt explained.
More budget
The survey shows that many of the AI challenges companies faced in previous years have decreased, with 52 per cent of companies reporting insufficient budget concerns have declined as firms begin to see the return on investments from AI.
The survey results show that over 70 per cent of respondents indicated they have seen a positive impact on their revenues from AI implementations. Additionally, the respondents have also seen a corresponding decrease in their annual costs due to AI.
“This is why we are seeing significant investment in AI – the initial proofs of concept and forays into AI have resulted in meaningful, measurable impacts,” Levitt said. “This is an exciting development because it indicates the industry is moving from just testing AI in a limited capacity to actually scaling AI applications into full production across the organisation.”
ESG
In the realm of ESG and sustainable finance, the report cites a notable transition from pilot systems to production capabilities, highlighting the growing maturity and integration of AI in sustainable finance initiatives.
According to Levitt, AI is a “winning formula” for applying to ESG research and reporting in the financial services industry, and using AI to extract and process all this necessary data is critical for producing accurate ESG reports and understanding a company’s progress.
“AI has made significant advancements over the past 18 months in its ability to handle both structured tabular data as well as unstructured data like voice, images, and PDFs,” Levitt said. “This AI capability to extract not just text, but also images, charts, and other content from various data sources is helping banks and other financial firms become more effective at monitoring and reporting on their ESG performance.”
