Workplace AI Users Report Correction Burden as Schools Struggle to Measure Impact

The Use.AI survey covered 8,316 adults who use AI in the workplace in the UK and European Union as well as the US, Canada, Australia and Latin America, TechRadar reported. It found that 48% said they had received AI-assisted work that required correction, and 43% said they had been forced to redo unsuitable AI work.

The review burden was not limited to fixing bad output. Thirty-four percent of respondents said they forwarded AI-produced or AI-assisted work without checking its contents. Thirty-nine percent said checking the AI's work took longer than creating their own version from scratch, and 41% said more reviewing and verifying was required after AI was brought into the productivity process.

The survey also recorded positive responses. Sixty-two percent said AI made it easier to create documents, memos and even code. But the same respondents reported problems at the next stage, when material had to be checked, corrected or redone. TechRadar reported that the findings raise questions about whether AI is delivering the productivity gains that have justified spending on tools, policies and personnel changes, as well as "AI-ready" marketing.

A separate figure pointed to gaps in oversight: 58% of respondents said there were no rules for checking AI-assisted work. TechRadar said that may account for the growing problem of AI-produced material being issued after a single prompt and distributed without any form of review.

In education, NPR reported that schools are experimenting with AI and that many are encountering the same problem: it is hard to measure whether the technology is good or bad for students. NPR did not give specific examples of school systems, test scores or other outcome measures in its report. The account described measurement difficulty as common among schools trying the technology.

The two reports address different settings but both center on what happens after AI tools are put to use. In the workplace survey, employees described time spent correcting AI output and inconsistent review rules. In schools, the obstacle NPR described was evaluating whether AI experiments produce measurable benefits or harms for students.