Concern about cybersecurity risks posed by AI surged among UK business leaders in March, rising seven points month-on-month (58% in March, up from 51% in February), representing the highest level of concern recorded to date.
The insight is revealed in Expleo’s AI Pulse sentiment tracker which combines levels of worry, excitement, trust and confidence into a single sentiment score on a 0–100 scale. The index also tracks AI-related cybersecurity sentiment which rose in March, putting the topic at the centre of this month’s data.
Jeff Hoyle, EVP and MD UK & North America at Expleo, said: “After a run of high-profile cyberattacks, it’s no surprise leaders are looking at AI and asking whether it makes things safer or risks widening the attack surface. But the risk here isn’t just data breaches. It’s sensitive data leaking into AI tools and weak controls around who can access what.”
Hoyle added: “Being aware of risks isn’t a bad thing. It’s a sign that leaders are taking security seriously as AI use grows. We’re already seeing organisations take action. Many are putting clearer policies in place, tightening controls and strengthening testing to make sure AI systems cannot be exploited.”
Increased worry around cybersecurity contrasts with the overall positive sentiment towards AI among the UK business community. The AI Pulse sentiment score for the UK in March is 64, the same figure as for February and a 10-point fall from its high of 74 in September 2025. This indicates that leaders can be excited about AI’s upside while still worrying about how it changes the security risk picture.
Hoyle commented: “This type of flattening off in positive sentiment combined with a greater awareness of risk is normal as new technology moves from early adoption to widespread implementation. It suggests we’re moving into a more stable phase where leaders are more focused on what it takes to successfully deploy AI.”
Trust in organisations to use AI ethically remains broadly stable (70%), around seven in ten UK business leaders expressing confidence in responsible deployment (70%) and a similar number saying they are excited about AI opportunities (71%). This indicates that sentiment around the technology’s potential remains strong, even as security concerns rise.
Hoyle concluded: “AI is still a bit of a gangly teenager, but it’s poised to grow up into something of a swan. This next phase for AI has got to be about delivering results that save time, improve quality and do so in a way that’s secure and controlled.”
