How to Check if the Text Is AI-Generated

Before getting started, all the suggested tools not only detect text generated from ChatGPT, they work for other AI-text generators like Rytr, Jasper, Bing Chat, etc. However, they are not completely accurate, so before coming to a decision, it is advisable to cross-check with different tools. Also, none of the tools guarantee accurate results that the content was AI-generated with nothing more than text. So do not use any of these tools to as the primary source for your decisions.

1. OpenAI’s AI Text Classifier

OpenAI itself has released a tool to check if the text is AI-generated, most importantly by ChatGPT. It is called text classifier and here are steps to use it:

  1. Open this link to open AI Text Classifier, scroll down to the Text section, and paste the text in the text field.
  2. Once done, click on the Submit button at the bottom.
  3. In a matter of seconds, Classifier will let you know whether the text is likely generated by AI or not. OpenAI’s AI Text Classifier does a good job of detecting AI-generated text. But, it is a bit limiting and responds by saying how likely the text was generated by AI. It does not share the likeliness in percentage like Copyscape does for example.

2. Content at Scale’s AI Detector

The classifier tool might be directly from the OpenAI itself, but it lacks many features such as highlighting text that is likely generated by AI, displaying the percentage of likelihood of content being generated by AI, etc. Whereas Content at Scale shows a human content score and highlights AI-generated sentences and paragraphs from the text you share. Remember that this percentage reflects probability and not the percent of text that is generated by AI. For example, if the tool says 85% human content score, that does not mean 15% is AI-generated. It means the system is 85% confident that the text you provided is created by humans.

  1. Open this link to open Content at Scale’s AI Detector. On the page, paste the text that you want to check in the text box.
  2. Scroll down and click on the Check for AI Content button.
  3. Now you should receive the Human Score in the left sidebar. This human score percent says how sure the system thinks it is that the content was created by humans. The higher the percentage, the more confident you can be that it is created by humans.
  4. On the right sidebar, you will see lines that the tool thinks are generated by AI.

3. GPTZero

Content At Scale’s AI detector highlight sentences that are probably generated by AI and the percentage it is confident. However, as we noted earlier, these results are not completely accurate but can give you estimated results. If you prefer to cross-check with another tool, we also recommend GPTZero.

  1. Open the GPTZero, scroll down, and paste the text in the text box. With GPTZero, you can also add documents. So if you have any word file or PDF, you can upload it to check and detect AI-generated content in that document.
  2. Now enable the checkmark beside I agree to terms of service and click on Get Results. button.
  3. Within a few seconds, the tool will show results on whether the text is generated by AI and even include any parts that are AI-generated.
  4. Scroll up and down to view highlighted text which means that it was probably generated by AI.

Limitations of AI Text Detectors

There are quite a few tools that may detect the text generated by AI. But so far, they all come with certain limitations like:

AI text detectors can be unreliable with short paragraphs. So make sure that the text is at least more than 1000 characters. Sometimes, the AI text detector can be completely unreliable and claim the text to be AI-generated even if it was written by humans. Though some language models can generate text in various languages, these AI text detectors only work in the English language for now. While all three AI text detectors can detect text generated by other language models, both AI text classifier and Content at Scale only work well when the text is generated by ChatGPT. They may not detect AI-generated text if they are edited by humans later.

  1. How Accurate Are These AI Text Detector Tools? The accuracy of these tools is low and cannot be considered the final decision-making criterion. Humans may create content that may seem to be generated by AI as the difference is nuanced. We recommend you crosscheck two or more tools to arrive at any sort of conclusion.
  2. Are Paid AI Text Detector Tools Any Better Than Free Tools? There are also some paid AI text detector tools such as Originality.ai. They offer more features like scanning multiple website pages, and stats about the article including the plagiarism checker, etc. But, when it comes to reliability, even such paid tools are not completely accurate.

Check if the Text Is AI-Generated From ChatGPT

You can use OpenAI’s own AI Text Classifier to check if the text is generated by AI tools like ChatGPT or manually created by humans. As these tools can be unreliable, we have included three other tools. It’s better to check on all three AI detector tools to reach a conclusion. All three tools are free and easy to use.