When AI Makes Office Work Look Better Than It Is

AI is very good at making work look finished. That might be the problem.

A recent Harvard Business Review piece argues that companies are walking into a strange new mess: employees are using AI to speed up everyday tasks, but the output can quietly get worse as it moves through the business. The writing looks clean. The report has the right shape. The resume sounds polished. But nobody is fully sure where the facts came from, who checked them, or whether the work still means anything.

The authors call this kind of low-effort AI output "workslop." It is a painfully good word. We have all seen some version of it by now: the email that says plenty but commits to nothing, the summary that sounds confident but misses the point, the document that feels like it was assembled instead of written.

The hiring example is the one that really sticks. A company uses AI to write the job ad. Applicants use AI to write their resumes. The company uses AI again to screen those resumes. Maybe another tool helps with interviews. By the end, what exactly is being judged? The person? The prompt? The software? It starts to feel less like hiring and more like two machines politely exchanging office paperwork.

That does not mean businesses should throw AI out. Some tasks are boring, repetitive, and honestly fine to automate. Nobody needs to lovingly handcraft every first draft of a meeting summary. But there is a difference between saving time and outsourcing judgment. The first is useful. The second gets expensive fast.

The simplest fix is also the least glamorous: companies need to know where information came from and who is responsible for checking it. If AI helped create a report, say so. If a human verified the numbers, say that too. And if nobody checked it, maybe do not treat it like fact just because it came in a neat paragraph.

AI can make office work faster. It can also make bad work travel faster. The trick is knowing which one is happening before the mess reaches the customer, the patient, the client, or the person sitting across the interview table.