Business leaders recognise the risks of generative AI (GenAI) but lack robust planning to deal with malfunctions, finds a new KPMG report.

Despite the promises of GenAI, business leaders have significant concerns about its implications for business performance, according to new research.

Respondents cited the inaccuracy of results including false information generated by large language models (LLMs) as the biggest concern when adopting GenAI (60%).

A total of 40 UK board members across different industries were polled with 53% worried about errors in the underlying data and information skewing an AI model’s outputs, with 50% noting cybersecurity as a growing concern.

Leanne Allen, head of AI at KPMG, said: “Given boards’ concerns, it is important that companies thoughtfully define a clear AI strategy rather than merely chase the next technological innovation.

“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 is a global organisation of independent professional services companies operating in 143 countries and territories, with more than 273,000 partners and employees around the world.

Just 30% of directors in the survey said that responsible GenAI usage guidelines have been published and communicated throughout their organisation to mitigate these potential issues.

KPMG believes this is an even greater concern after a recent poll by the audit, tax and advisory company revealed that 42% of UK adults have entered work-related information into GenAI tools.

Another concern for board members was algorithmic bias, meaning repeated errors in a computer system that create outcomes privileging one category or action over another.

Only 8% of respondents indicated that their organisation had a process in place to measure AI bias.

A further 23% of businesses were found to have implemented GenAI training as well as measures to mitigate other risks associated with the adoption of AI.