Understanding AI Humanisers
An AI humaniser is a tool designed to take text generated by large language models (LLMs) like ChatGPT or Claude and revise it to sound more like a person wrote it. These tools work by analysing the input for common AI patterns—such as repetitive phrasing,
predictable sentence structures, and a formal tone—and then rewriting it. The goal is to improve readability and flow, making the text more engaging and, in many cases, attempting to bypass AI detection software. They swap synonyms, vary sentence length, and adjust the overall rhythm to mimic a more natural, human voice.
The Core Dilemma: Authorship
The use of these tools raises a critical question: if an AI drafts the text and another AI refines it, who is the author? Current legal frameworks in most major jurisdictions, including the US and EU, do not recognize AI as a legal author; copyright requires human creativity. When AI is used merely as a tool, similar to a word processor, the human operator can typically claim authorship. However, when the AI's contribution becomes substantial, the lines blur. This creates a significant ethical gray area for professionals, students, and creators whose work depends on originality and personal accountability. Submitting heavily modified AI text as your own can be seen as a breach of academic or professional integrity.
The Art of Disclosure
Given the rise of AI, transparency has become paramount for maintaining trust. The question is no longer if you should disclose AI use, but when and how. For minor uses, like correcting grammar or spelling, disclosure is rarely necessary. However, for any substantive use—such as drafting sections of an article, generating ideas that form the core of your work, or using a humaniser to heavily edit text—disclosure is best practice. Think of it not as a confession, but as routine methodological reporting, similar to citing software used in research. A simple line in an acknowledgements section or a footnote stating that AI was used to assist in drafting or editing can build credibility rather than diminish it.
Practical Choices for Authentic Writing
Instead of relying on an AI humaniser to fix robotic text, a better approach is to change how you use AI in the first place. Use generative AI as a brainstorming partner, a research assistant, or a tool for creating a rough first draft. Then, take over. The most effective way to “humanise” AI content is through your own critical thinking, unique voice, and personal insights. Inject your own anecdotes, challenge the AI's assumptions, and restructure the content to fit your narrative. This human-in-the-loop process not only ensures higher quality and authenticity but also keeps you firmly in the author's seat. The goal shouldn't be to make AI text undetectable; it should be to create genuinely valuable content.
The Risk of an Arms Race
It's important to recognise the limitations and risks of relying on humanisers. AI detection tools are constantly evolving, and what is “undetectable” today may be easily flagged tomorrow. Furthermore, these tools can sometimes introduce awkward phrasing or errors, degrading the quality of the text in their attempt to evade detection. This creates a pointless arms race between humanisers and detectors. Focusing on this cat-and-mouse game distracts from the ultimate goal of any writer: to communicate clearly and build a relationship with an audience based on trust and quality. Research shows that while audiences want transparency, they may also lose trust in content they know is AI-generated, reinforcing the need for genuine human oversight and authorship.
