Rushing out AI likely to backfire, report

Hype is fuelling fear and redundancies

Businesses investing heavily in AI may be inadvertently creating cultures of fear and uncertainty, according to new research from tech consultancy The Adaptavist Group.

The study, part of Adaptavist’s Digital Etiquette: Unlocking the AI Gates report, surveyed 900 professionals responsible for implementing AI across the UK, US, Canada and Germany. It uncovered a stark divide between organisations whose AI claims are perceived as over-inflated, dubbed AI ‘sceptics’, and those taking a more measured approach, the ‘realists’.

Nearly 42% of AI leaders said their company’s AI promises were exaggerated, compared with 36% who felt claims were realistic. The difference in perception is significant: sceptic-led organisations report higher levels of fear, pressure and potential risk to both employees and customers. Among sceptics, 65% believed their organisation’s stance on AI put customers at risk of financial, psychological or physical harm, compared with just 9% of realists. Job insecurity was also markedly higher, with 67% worried AI could threaten roles, compared with just 10% in realist organisations.

Yet investment levels were broadly similar. Over a third of sceptics’ organisations had spent between £1 million and £10 million on AI in the past year, with 8% investing over £10 million. Realists’ organisations were only slightly behind, suggesting that perception rather than investment drives the disparity in outcomes.

The report highlights how top-down implementation can backfire. In sceptic-led companies, 84% of leaders said their teams were encouraged to use AI because they felt they should, not because it delivered measurable value, compared with 58% of realists. Half of sceptic organisations tied AI use to performance KPIs, while 66% discouraged experimentation, stifling curiosity and innovation.

Training disparities were also stark, with 59% of sceptic leaders reporting no formal AI training compared with just 16% of realists. This gap contributed to anxiety, with 42% of sceptic leaders hiding their AI use for fear of repercussions, while 48% worried about being incorrectly accused of using it.

By contrast, organisations taking a more honest approach to AI adoption reported far better results. Leaders in realist organisations reported improvements in work quality, time efficiency, and output far more frequently than sceptics. They also raised far fewer concerns about ethics, accuracy, hallucinations and bias, reflecting the stronger training and support they had received.

Rapid AI tool proliferation appears to be widening the divide, 74% percent of sceptic leaders said too many AI tools were being introduced too quickly, compared with 24% of realists. Two-thirds of sceptics expect their AI usage to decline over the next year, compared with just 9% of realists.

Jon Mort, CTO at The Adaptavist Group, said the research shows that perception and implementation matter more than investment alone. “Organisations treating AI simply as a replacement for existing tasks, without considering training or system-wide integration, are set up to fail,” he said. “Measured rollout, experimentation and supportive cultures are critical to unlocking AI’s potential.”