Science & Tech · AI · 4 hrs ago
Few UK workers check AI answers before using them, survey finds
A survey of 2,000 UK workers found many rely on AI at work for advice instead of asking colleagues or managers.
Only 24% check AI answers before acting on them.
The survey, conducted by Censuswide for learning company O’Reilly, found that 39% had acted on inaccurate or unsuitable AI responses.
Training is limited: 16% received AI training related to their role, while 35% said they had received no AI training or guidance in the past year.
The findings point to a gap between fast AI adoption and workplace training.
O’Reilly says employers should provide role-specific training and encourage workers to verify AI answers; the survey does not specify what organizations will do next.
A survey of 2,000 UK workers found that only 24% check AI-generated answers before acting on them.
The research found that 39% had acted on AI responses that proved inaccurate or unsuitable.
Only 16% said they had received AI training related to their role, while 35% said they had received no AI guidance or training.
The survey was conducted by Censuswide on behalf of learning company O’Reilly.
O’Reilly’s Alexia Pedersen said organizations should address training and verification rather than slow AI adoption.
- Who
- The survey covered 2,000 UK workers.
- What
- It found that few workers verify AI-generated answers before using them.
- When
- Not stated.
- Where
- The survey covered workers in the UK.
- Why
- Not stated.
This story does not have two clearly opposing sides.
This story does not have a timeline yet.
- Workers surveyed
- 2,000 UK workers
- Verify AI answers
- 24%
- Acted on inaccurate or unsuitable AI responses
- 39%
- Role-related AI training
- 16%
- No AI guidance or training
- 35%
What people are saying
We're seeing a workforce that has fully embraced AI as a first port of call, but the safety net hasn't caught up with the speed of adoption,
People are moving faster than their organizations' training and governance.
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.
This is a training and trust problem, not an AI problem.