Survey finds 39% of workers say checking AI output takes longer than doing the task
A global study reveals that 39% of employees find verifying AI-generated content takes longer than creating it from scratch.
Key points
- % of respondents said checking AI work takes longer than doing the task themselves.
- % reported receiving AI-assisted work that required substantial correction.
- % forward AI-produced material without verifying its accuracy or suitability.
A new survey conducted by Use.AI indicates that workplace artificial intelligence often requires substantial revision. The study of 8,316 adults across multiple regions found that 39% of respondents said checking the AI’s effort took longer than creating their own version from scratch.
The research highlights a discrepancy between the ease of generation and the difficulty of verification. While 62% of participants noted that AI made creating documents, memos, and code easier, 48% stated they had received AI-assisted work that required correction. Additionally, 43% said they were forced to redo unsuitable AI work entirely.
Productivity Impact
The data suggests that the "fire and forget" approach to AI is flawed. Thirty-nine percent of respondents reported that checking the AI’s effort took longer than creating their own version from scratch. Furthermore, 41% highlighted how much more reviewing and verifying AI-produced material was required since AI was brought into the productivity process.
This additional oversight creates a bottleneck. The study found that while AI can produce documents employees might not have created themselves, the technology creates more problems for almost 48% of users. The need for substantial revision undermines the initial speed gains provided by the tools.
Global Scope
The Use.AI study surveyed adults in the UK, European Union, US, Canada, Australia, and Latin America. The findings cover these diverse markets, indicating widespread issues with current AI implementations in professional settings.
A concerning 34% of respondents admitted to forwarding AI-produced or assisted work without checking the contents. This practice can lead to the distribution of unsuitable or unreadable material. The lack of verification steps exacerbates the time spent on corrections downstream in the workflow.
What should you do?
Organizations should evaluate their AI integration strategies. The study implies that without proper oversight, AI tools may reduce overall efficiency. Companies need to balance the speed of generation with the necessity of accurate, verified output.
What to do and how to stay safe: Use.AI
- Implement mandatory review protocols for all AI-generated content before distribution.
- Train staff to identify common errors and hallucinations in AI-assisted work.
- Measure the total time spent on generation plus verification versus manual creation.
- Establish clear guidelines on when AI assistance is appropriate for specific tasks.
Step-by-step guide: Prompt Injection Warning Signs: How to Spot AI Manipulation
General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
How many people were surveyed for this study?
The Use.AI study surveyed 8,316 adults who use AI in the workplace across several countries.
What percentage of workers forward AI work without checking it?
According to the report, 34% of respondents forward AI-produced or assisted work without checking the contents.
Does AI save time for most workers?
No, 39% said checking AI output takes longer than doing the task themselves, and 48% found it created more problems.
