Over half (55 per cent) of UK marketers say their organisation has made a significant business decision based on AI-generated insight which later turned out to be wrong or misleading, according to new data from GWI, a global insights and intelligence company.
The research, which surveyed 150 marketers across the UK, explores how marketers’ confidence in AI-generated insight is outpacing the discipline to verify it, a gap that’s already producing significant costs for businesses across the globe.
The cost of acting on inaccurate or misleading insights is already measurable. Almost a third (29 per cent) of marketers say it has led to financial losses for their organisation, while 35 per cent report customer complaints and 29 per cent reputational damage.
“Marketing teams are under pressure to turn insight into action quickly,” said Jason Mander, Chief Insights Officer at GWI. “But when an AI-generated answer informs a campaign or strategic decision, its accuracy matters even more than its speed. These findings show the business cost of acting on incorrect insights.”
The consequences also extend to day-to-day work and campaign results. Almost half (47 per cent) say inaccurate AI outputs have forced teams to redo work, while 54 per cent say it’s delayed decision making, and 17 per cent say it has damaged campaign performance.
The findings also raise questions about how consistently teams check AI-generated insights before they act on them. While the majority (91 per cent) believe the risk of making business decisions based on inaccurate AI-generated insights is significant, only 29 per cent say they verify AI-generated insights “most of the time.”
And AI isn’t just getting things wrong, it’s making old information look brand new. More than four fifths (84 per cent) of marketers agree that AI makes it easier for outdated information to influence decisions, while 79 per cent have accidentally used data that they later realised was out of date.
“AI is only as reliable as the data behind it,” Mander adds. “Verifcation means knowing where an insight comes from, how the underlying data was collected and whether it’s still up to date. These checks need to be built into everyday workflows, so that when an insight informs a campaign or business decision, there’s strong evidence behind it.”
These findings offer a preview snapshot from GWI’s upcoming report, Artificial Confidence: The AI Trust Gap, launching in November, which explores how businesses are using AI to understand their audiences, the risks of relying on unverified insights, and the role of trusted human data in supporting decision-making.
Share