Bad news for students using ChatGPT for homework: They will be detected

OpenAI, the developer of ChatGPT, is working on a new watermark tool that can instantly detect text generated by the AI. This development means that students who rely on ChatGPT to complete their assignments could be identified.

OpenAI has introduced ChatGPT as a powerful language model capable of providing natural responses to a wide array of questions, much like a personal tutor. However, to prevent misuse, the company is now taking steps to ensure academic integrity.

The new tool being developed by OpenAI aims to identify text created by ChatGPT, a feature that will likely be welcomed by academic institutions and educators. The increasing trend of students using AI tools like ChatGPT for their homework has become a significant issue in the academic world. This detection tool is expected to help maintain academic honesty by identifying such instances.

OpenAI's Uncertainty

Despite the promise of this watermarking technology, OpenAI remains cautious about its deployment. The company acknowledges the potential risks associated with the tool, including the possibility of misuse by malicious individuals and the likelihood of incorrect results in non-English languages.

How the System Will Work

The new detection method is expected to be much more effective than current techniques. The tool will specifically target text generated by ChatGPT, subtly altering the choice of words and adding an "invisible watermark" to the final text. This hidden marker will allow various tools to easily detect whether a piece of text was created by ChatGPT.

In summary, OpenAI's watermark tool aims to curb the misuse of ChatGPT in academic settings, ensuring that students' work remains genuine and maintaining the integrity of educational assessments.