New survey: 76% of top financial services firms are envisaging or already using AI or machine learning

More than three quarters of financial services firms surveyed are either considering or are already using artificial intelligence (AI) or machine learning (ML) solutions, according to a new global report by international law firm Norton Rose Fulbright.

The firm’s new The Power of AI: How is artificial intelligence transforming financing services? report features responses from a cross-section of financial services clients* surveyed around the world.

While only six percent of respondents said they are at the point of moving towards large scale adoption or development of AI/ML solutions, many confirmed they are in the early stages of assessing or developing the tools’ potential across at least some areas of their business. Less than a quarter (24 percent) of respondents said that they do not currently use AI/ML at all.

A total of 94 percent say that they will (71 percent) or may (23 percent) use AI/ML solutions more in the next three years. A sizeable 87 percent of respondents said that they are already using, have confirmed plans to use, or are already evaluating use of large language models (LLMs) such as ChatGPT or LaMDA.

The survey reveals that use cases for AI/ML are quite varied, with many currently involving automation – for example, document generation and IT security – rather than automated decision-making.

Barriers to adopting AI/ML among surveyed financial services firms are also varied, with a lack of expertise (23 percent) and limitations around the availability of data and/or the right to use such data (19 percent) both mentioned as key factors. The risks that financial services firms are most concerned about include training data and data protection compliance (28 percent) and biased outcomes (24 percent).

Marcus Evans, Partner at Norton Rose Fulbright LLP, said:

“The efficacy of machine learning models depends on access to sufficient amounts of training data. This can be a challenge to source internally or externally. Synthetic data sources and extensive training of the system might be required. It is therefore no surprise that 19 percent of financial services firms responding to our survey take the view that data limitations are a key barrier for adoption.”

Many financial services firms appear to still be in the planning stage, with relatively few respondents (15 percent) saying they are currently discussing their plans to adopt AI/ML with regulators.

Hannah Meakin, Partner at Norton Rose Fulbright LLP, said:

“In the UK, to take an example, the Prudential Regulation Authority and the Financial Conduct Authority will want to see firms putting in place governance arrangements around all stages in the AI lifecycle. Robust governance is likely to be one of the key means through which firms can give the regulator comfort about their evolving use of AI.”

Norton Rose Fulbright advises many of the world’s leading financial institutions and FinTechs on the legal, regulatory and commercial issues raised by the impact of technology on their businesses. Operating in more than 50 cities across the world, the global team comprises experienced, technology and regulatory legal professionals, with specialised industry experience from across all of the firm’s major practice areas.

*The report features responses from more than 50 of the world’s top financial services firms, who are clients of Norton Rose Fulbright.