Educational institutions are facing a growing challenge as artificial intelligence tools become increasingly capable of helping students complete assignments and research projects with minimal effort. At the same time, recent reports suggest that AI-content detection systems are struggling to keep pace with new methods designed to evade detection.

According to a report by the New York Times, a wave of applications promoted on platforms such as TikTok and YouTube now enables students to conceal their use of AI-generated content more effectively. Some of these tools rewrite text produced by AI models, making it appear more natural and closer to human writing styles, significantly reducing the likelihood of being flagged by traditional detection software.

Other applications use a technique known as “progressive writing,” where text is gradually inserted into documents over several hours. These tools can also introduce artificial edits and spelling mistakes to mimic a genuine human writing process rather than pasting a completed AI-generated text all at once.

Detection Companies Face Criticism

The report also highlighted concerns that some companies developing AI-detection technologies may be contributing to the problem. One example is GPTZero, a platform designed to identify AI-generated content. The company faced criticism after it emerged that a marketer associated with the platform promoted its services through student-focused TikTok accounts, explaining how users could test assignments before submission.

Similarly, Grammarly has drawn scrutiny for offering AI-detection tools alongside text-generation and paraphrasing features, raising questions about the apparent contradiction between detecting AI-generated content and providing tools that can help reshape it.

High Failure Rates Raise Concerns

Concerns have intensified following recent academic research. Studies conducted by researchers at the University of Florida found that some leading AI-content detection systems failed to identify modified AI-generated text in up to 99.6% of cases.

The research also revealed that relatively minor changes in wording, sentence structure, or writing style could be enough to bypass many of today's detection tools.

Is Banning AI Still Practical?

Given these challenges, some experts argue that banning AI tools in schools is no longer a practical solution, particularly because proving AI usage with certainty has become increasingly difficult.

Many educators believe the focus should shift from restricting technology to teaching students how to use AI responsibly and ethically. They argue that AI tools are likely to become an integral part of future workplaces, making digital literacy and responsible usage more important than outright prohibition.