When OpenAI launched ChatGPT in late 2022, teachers were concerned that students would use it to cheat. Since then, various AI detection tools have seen the light of day, with varying percentages of accuracy.
OpenAI has now developed a method to detect the use of ChatGPT with 99.9 percent accuracy by embedding an undetectable watermark in the text. Despite being ready for nearly a year, the tool’s rollout has been delayed by internal debate and concerns that it disproportionately affects non-native English speakers.
While some believe the tool could benefit educational attainment, others worry it could turn users away. The tool’s release date is still open for the time being.
Source: Android Authority
