A multinational survey of nearly 10,000 workers finds that many employees appear more experienced at work because of their use of artificial intelligence (AI). The study, conducted by Use.AI and reported by Euronews, covered respondents in the United States, the United Kingdom, Canada, the European Union and Latin America.
The research provides several concrete figures:
- 64% of respondents have used AI to complete a task they could not have accomplished on their own.
- 43% reported taking on responsibilities with AI assistance that they felt they were not yet fully ready for.
- More than half of those surveyed said AI made them appear more experienced at their workplace.
These findings pose challenges for employers: finished outputs increasingly fail to reveal what an individual can perform unaided. According to the survey, 35% of respondents believe they would struggle with certain parts of their current job without AI, and about a quarter fear their employer values their abilities more highly than their unaided competence.
Transparency and disclosure: when should AI use be reported?
The study also highlights limited transparency in practice: 39% of respondents submitted work created with AI without disclosing it, and 30% accepted praise for work to which AI made a significant contribution. Additionally, 19% said AI-assisted work contributed to their promotion.
Use.AI's researchers argue that it would be unrealistic to require disclosure of every single AI interaction. However, they say transparency is justified when the technology materially shaped the content of an analysis, presentation, recommendation, program code or other important work.
Performance evaluation and new competency criteria
The spread of AI may reshape performance reviews. Because the final product increasingly conceals the human knowledge behind it, employers may need to assess other abilities, such as:
- whether the worker can explain their own reasoning;
- whether they can recognise AI errors;
- whether they can make decisions with incomplete information;
- how they respond when a system provides an incorrect answer.
Particularly valuable may be the ability to clearly define a problem, ask the right questions and justify decisions. According to Use.AI's leader, the situation becomes problematic when someone can produce a convincing answer with AI but cannot recognise if it is wrong, cannot explain why it is correct, or cannot make appropriate professional decisions without the technology.
In sum, the study suggests AI can significantly extend what an employee can achieve, but organisations and employers will need new methods to assess true independent capability and to ensure transparency and appropriate allocation of responsibility.
