OpenAI has recently addressed the issue of whether their AI detectors are capable of detecting AI-generated content on educational websites, specifically for teachers and professors. They admit that their system is unable to determine whether a student’s homework or written work has been done using AI or not. In response to the question “Do AI detectors work?” posed to OpenAI, their straightforward answer is “In short, no.” They further elaborate that while many companies, including OpenAI, have attempted to develop reliable detection systems, none of these systems are currently deemed trustworthy enough. OpenAI also warns users that their ChatGPT system does not possess the capability to be aware of the AI-generated nature of the content fed into it. Therefore, if users encounter responses from ChatGPT asking if the content was written by AI, they should remember that it is merely a random and nonspecific question, and does not indicate an ability to detect AI-generated content.
In conclusion, OpenAI acknowledges the limitations of AI detectors in recognizing AI-generated content and advises caution when relying on their ChatGPT system.
TLDR: OpenAI states that their AI detectors are unable to detect AI-generated content accurately and emphasizes that their ChatGPT system is not equipped with knowledge of the AI origins of the content it generates. Users should exercise caution when interpreting responses from ChatGPT.