原文开头
For many browbeaten office workers, liberation—or, at least, a better-mannered boss—may be nigh. A growing category of software is using AI to detect workplace bullying, raising the prospect that toxic managers could soon have fewer places to hide. By Jo Constantz While forms of employee-surveillance software have been around for decades, the advent of large language models marks an inflection point. Earlier generations of surveillance software relied on keywords (for example, four-letter words) and rudimentary natural- language-processing techniques to detect abusive language and behavior. With more sophisticated artificial intelligence systems, employers can take context, tone and recurring patterns into account across thousands of conversations to help identify problematic managers far earlier. …
摘自《彭博商业周刊》(Bloomberg Businessweek)2026年7月刊。仅引用开头一小段供了解文章,版权归原刊所有,全文请阅读原刊。