AI cheating crisis deepens as 95% of US students use chatbots

Artificial intelligence is intensifying concerns about academic dishonesty in the United States, with a study reportedly finding that 95 per cent of students use AI chatbots to write their assignments.

The growing use of AI tools in education has raised concerns among teachers and researchers about whether students are genuinely learning or simply creating the impression that they understand their subjects. Educators fear that students may be completing academic tasks without developing the critical thinking, writing and problem-solving skills those assignments are designed to build.

According to the reported findings, AI chatbots are increasingly being used to generate written assignments. While these tools can help students understand complex topics and organise their ideas, their use to produce work that students submit as their own has created new challenges for educational institutions.

The problem has become more difficult to address with the emergence of AI programmes designed to make machine-generated text sound more natural and human-like. These tools can alter the style and wording of AI-produced content, making it harder for teachers to distinguish between work written by students and text generated by chatbots.

Researchers have described the situation as a growing contest between AI detection systems and tools designed to evade detection. As AI-generated writing becomes more sophisticated, educators face increasing difficulties in identifying academic misconduct through conventional methods.

The debate has also highlighted the limitations of AI-detection software. Such systems may struggle to identify rewritten or humanised AI content, while their results can also be unreliable in some cases. This has raised questions about how schools and universities should assess students’ work fairly.

Teachers are increasingly concerned that the widespread use of AI could weaken the learning process if students rely on chatbots instead of developing their own understanding. Completing an assignment with AI assistance may produce an impressive result without ensuring that the student has mastered the subject.

The findings have renewed calls for educational institutions to reconsider assessment methods and establish clearer rules governing the use of AI. Possible approaches include greater emphasis on in-class writing, oral examinations, project discussions and assignments that require students to explain their reasoning.

AI tools can still play a useful role in education when used responsibly. However, educators face the challenge of ensuring that students use the technology to support learning rather than replace it.

Note: The reported 95 per cent figure is presented as stated in the supplied report. The original study, its sample size and its methodology were not provided, so the statistic could not be independently verified.