A Wikipedia Group Made a Guide to Detect AI Writing. Now a Plug-In Uses It to ‘Humanize’ Chatbots
The Humanizer: A Plug-In That Tries to Make Chatbots Sound More Human
A group of Wikipedia editors has been working on a guide to detect AI writing, and now a plug-in uses that guide to "humanize" chatbots. The plug-in, called Humanizer, was released by tech entrepreneur Siqi Chen and is designed to work with Anthropic's Claude Code AI assistant. But does it actually work, and what does it mean for the future of AI writing?
The Wikipedia Guide to Detecting AI Writing
The Wikipedia guide, created by WikiProject AI Cleanup, is a list of 24 language and formatting patterns that are commonly associated with AI-generated writing. The group of editors, led by French Wikipedia editor Ilyas Lebleu, has been working on the project since late 2023 and has tagged over 500 articles for review. In August 2025, they published a formal list of the patterns they kept seeing.
Some of the patterns include:
- Using phrases like "marking a pivotal moment" or "stands as a testament to"
- Writing like tourism brochures, calling views "breathtaking" and describing towns as "nestled within" scenic regions
- Tacking "-ing" phrases onto the end of sentences to sound analytical
How the Humanizer Plug-In Works
The Humanizer plug-in uses the Wikipedia guide to instruct the Claude Code AI assistant to avoid using these patterns and instead write in a more human-like way. The plug-in is a "skill file" that adds a list of written instructions to the prompt fed into the large language model that powers the assistant.
The skill file includes instructions like:
- "Have opinions. Don't just report facts—react to them. 'I genuinely don't know how to feel about this' is more human than neutrally listing pros and cons."
- "Replace inflated language with plain facts."
Does the Humanizer Plug-In Actually Work?
In limited testing, the Humanizer plug-in made the AI agent's output sound less precise and more casual. However, it's not clear whether this is actually an improvement or just a change in style. The plug-in may also have some drawbacks, such as:
- Not improving factuality
- Potentially harming coding ability
Why AI Writing Detection Fails
Even with a confident set of rules crafted by Wikipedia editors, AI writing detectors don't work reliably. There is nothing inherently unique about human writing that reliably differentiates it from LLM writing. One reason is that even though most AI language models tend toward certain types of language, they can also be prompted to avoid them.
The Future of AI Writing
The Humanizer plug-in and the Wikipedia guide to detecting AI writing highlight the challenges of creating AI writing that sounds like human writing. While the plug-in may be able to make chatbots sound more human, it's not clear whether this is actually an improvement or just a change in style.
Ultimately, the future of AI writing will depend on the ability to create AI systems that can understand and replicate the complexities of human language. This will require advances in natural language processing, machine learning, and cognitive science.
Conclusion
The Humanizer plug-in and the Wikipedia guide to detecting AI writing are just two examples of the many challenges and opportunities in the field of AI writing. As AI systems become more advanced and widespread, it's likely that we will see more attempts to create AI writing that sounds like human writing. However, it's also likely that we will see more attempts to detect and prevent AI-generated writing from being passed off as human writing.
Ultimately, the future of AI writing will depend on the ability to create AI systems that can understand and replicate the complexities of human language. This will require advances in natural language processing, machine learning, and cognitive science.
