AI adoption seeing only ‘modest growth’ across organisations, survey reveals

Only one in 10 organisations report having implemented large-scale AI implementation despite its rate of predicted market growth, a recent survey finds.

The Nash Squared Digital Leadership Report found that while artificial intelligence (AI) has grabbed the attention of digital leaders globally, the rate of uptake has remained relatively modest over the past five years.

Only 10 per cent of organisations have implemented large-scale projects involving AI in the past year.

Large-scale projects relating to the cloud, on the other hand, have doubled, with the proportion of organisations undertaking large-scale cloud implementations jumping from 34 per cent to 68 per cent over a five-year period.

However, the survey report said that generative AI might be the trigger that sees it follow a path similar to cloud investment.

“While generative AI has been in development for a while, its recent spurt of growth has been the result of massive leaps in hardware processing power and clever refinement of algorithms,” the report said.

“No technology has arrived, been adopted or had such an immediate effect so quickly. This rapid arrival has left digital leaders thinking hard. Some see it as no more than a sophisticated virtual assistant; others see it as a more fundamental shift in what technology can do for, and maybe in place of, humans.”

Data privacy concerns

More than a third of digital leaders are concerned about data privacy as an issue for implementing AI.

“While AI itself is not fundamentally insecure, when new technology arrives it tends to open up new areas of exposure,” the report said.

“Among all of this highly sophisticated technology, it’s sometimes easy to forget that cutting and pasting a confidential business strategy into a generative AI tool to summarise it or change the tone is a potential data breach. Education and awareness will be key to mitigating against this.”

Around a quarter are also concerned about eliminating hallucinations in the data, where the large language model (LLM) generates false information.

“Hallucinations can range from minor inconsistencies to completely fabricated or contradictory information,” it said.

Most digital leaders feel unprepared for generative AI

The survey also found that only 15 per cent of digital leaders feel “very” or “extremely well prepared” for the demands of generative AI, and 42 per cent said they feel unprepared.

Only one in two respondents in the survey have an AI policy in place in their organisation.

“More than a third have no plans to even attempt a policy at this time,” the report said.

This article was originally featured in Accounting Times.