AI hallucinations, accuracy still top concerns for UK tech leaders as adoption continues
Generative AI guidance and education is thin, a new study shows, creating hesitancy among IT leaders
Business leaders are wary of generative AI, according to new research from KPMG, with many citing major concerns about its impact on business performance.
Six-in-ten tech leaders told KPMG that the accuracy of results and the potential for hallucinations are their biggest concern when adopting generative AI tools.
Boards are also worried about errors in the underlying data and information skewing the model’s outputs ,with 53% expressing concerns. Meanwhile, half said they were worried about problems related to cybersecurity.
Notably, only three-in-ten directors said responsible generative AI usage guidelines have been published and communicated throughout their organization to mitigate these potential issues.
This, the consultancy said, is even more concerning given that 42% of UK adults told a separate KPMG poll that they had entered work-related information into generative AI tools.
Algorithmic bias was identified as another major worry by board members, with 43% highlighting it was a serious issue. Despite this, just 8% of the survey respondents indicated that their organization has processes in place to measure it.
"Given boards’ concerns, it’s important that companies thoughtfully define a clear AI strategy rather than merely chase the next technological innovation," said Leanne Allen, Head of AI at KPMG UK.
Get the ITPro. daily newsletter
Receive our latest news, industry updates, featured resources and more. Sign up today to receive our FREE report on AI cyber crime & security - newly updated for 2024.
"This strategy should balance the value, cost, and risk associated with AI use cases. This strategic equilibrium is crucial for both progress and stakeholder trust."
KPMG’s research did highlight positive findings, however. Almost one-quarters of firms said they now have dedicated generative AI training in place to mitigate potential risks, or at least in the process of developing educational materials.
That said, nearly seven-in-ten business leaders noted that board education on generative AI in their organization today is primarily ad hoc or self-taught by individual directors.
Nearly half said that board education involved management presentations, with a similar number citing third-party training. Only one-in-twenty boards have, or are actively recruiting for, generative AI expertise.
In a report earlier this summer, enterprise software and cloud procurement firm SoftwareOne found that fewer than half of UK companies had the right skills to make the most of AI, with a similar proportion saying they were struggling to find AI-skilled employees.
"Rapid advancements in AI and generative AI offer exciting prospects for companies worldwide, but organizations are sitting on a ticking talent time bomb if they don’t upskill and retrain their workforces now to fulfill the potential of AI," said SoftwareOne CEO Brian Duffy.
Last year, the UK government launched new guidance on AI training aimed at helping organizations to make sure that their workers can safely make the most of the technology.
Emma Woollacott is a freelance journalist writing for publications including the BBC, Private Eye, Forbes, Raconteur and specialist technology titles.