On the earth of unreliable AI detection, Undetectable AI rose as an unlikely hero that protects you from getting falsely accused of utilizing LLMs. However, to satisfy its guarantees, it has a darkish however open secret:
It deliberately writes grammatically incorrect outputs to keep away from AI detection.
There’s nothing we might do about that…
![](https://terracyborg.com/wp-content/uploads/2024/06/image3-9-840x473.png)