Researchers from the University of Reading in the UK deceived their own professors by secretly sending them AI-generated exam answers. These answers received higher grades than those of real students.

As part of the project, the researchers created fake student accounts. Through these accounts, they sent unedited exam answers generated by ChatGPT-4 to the professors. This took place during online home assessments of undergraduate courses. The university's professors only suspected one out of 33 papers. The rest of the AI-generated answers received higher grades than the average grades of real students.

According to the researchers, their findings indicate that AI, particularly ChatGPT, is now passing the "Turing test" – a test created in 1950 to determine a machine's ability to exhibit intelligent behavior. This means they can go unnoticed even by experienced professors.

The authors note that AI's ability to demonstrate more abstract reasoning will improve, while the ability to detect it will diminish. Therefore, the issue of academic integrity will deepen.

Source: https://life.pravda.com.ua/society/u-britaniji-profesori-vishu-ne-rozpiznali-zgenerovani-shi-vidpovidi-na-ispit-302368/