UK workers rarely verify AI answers; 39% act on inaccurate guidance

Survey reveals low verification and high reliance on AI

A survey conducted by Censuswide on behalf of learning company O’Reilly Media found that only 24% of UK workers verify AI-generated answers before using them in their work. Nearly 39% of respondents admitted to acting on AI guidance that later proved inaccurate or unsuitable, according to the findings published by TechRadar.

The survey highlights a growing preference for AI chatbots over colleagues or managers for work-related advice, despite the low rates of verification. This trend suggests a shift in workplace trust dynamics, with employees increasingly relying on AI tools without oversight.

Training gaps exacerbate risks

The survey also exposed significant gaps in AI training. Only 16% of UK workers reported receiving role-specific AI training, while 35% said their employers provided no AI guidance or training at all.

Alexia Pedersen, SVP International at O’Reilly, said the issue is not AI itself but a "training and trust problem." She added, "The answer isn’t to slow down AI adoption; it’s to close the gap with the right training, verification habits, and trusted expertise to back it up."

Link to broader AI governance challenges

These findings align with earlier Lazyfounder reporting on UK AI investment trends, which revealed that governance gaps and poor oversight are costing businesses billions in failed AI projects. While UK firms have invested heavily in AI, many struggle to measure its value or achieve a positive return, often due to fragmented responsibility and lack of verification protocols.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

The survey’s findings are a warning for founders and operators: AI adoption without guardrails can lead to costly mistakes, particularly in fast-scaling startups where errors can compound quickly. The preference for AI chatbots over human advice also signals a shift in workplace trust, which could reshape team dynamics if left unchecked.

For investors, the results reinforce concerns about AI governance. Without structured training and verification, AI investments risk underdelivering, mirroring the broader UK market struggles with AI ROI. Startups that proactively address these gaps—through training, oversight, and clear protocols—may avoid the pitfalls seen in larger organizations.

Key takeaways

• AI adoption without training and verification risks costly errors, particularly for startups where operational mistakes can scale quickly.
• Founders may need to invest in structured AI training programs to bridge gaps identified in the survey, as reliance on AI without oversight could erode workplace trust.
• The findings mirror broader governance challenges highlighted in UK AI investment trends, where lack of oversight correlates with poor ROI on AI projects.

Related on Lazyfounder

• UK AI Investment Soars, But Governance Gaps Erode Billions in Potential ROI
• More ai news