Source
Forbes
Summary
The article reveals a growing trend: students are using “AI humaniser” tools to mask signatures of ChatGPT-generated essays so they pass AI detectors. These humanisers tweak syntax, phrasing, rhythm and lexical choices to reduce detection risk. The practice raises serious concerns: it not only undermines efforts to preserve academic integrity, but also escalates the arms race between detection and evasion. Educators warn that when students outsource not only content but also disguise, distinguishing genuine work becomes even harder.
Key Points
- AI humaniser apps are designed to rewrite AI output so it appears more human and evade detectors.
- The tools adjust stylistic features—such as sentence variety, tone, and lexical choices—to reduce red flags.
- Use of these tools amplifies the challenge for educators trying to detect AI misuse.
- This escalates a detection-evasion arms race: detectors get better, humanisers evolve.
- The phenomenon underlines the urgency of redesigning assessment and emphasising process, not just output.
Keywords
URL
Summary generated by ChatGPT 5
