My Response to Those Who Claim AI Detectors Are Useless

For those claiming that “it’s a waste of time to try to detect student (AI) use,” or “AI detectors are useless,” I’d caution that, while you might be right, such extreme positions might also be found wanting as the technology evolves.

In other words, for any of you who are thinking about making a similar claim, I suggest first taking into account the context and key factors mentioned below.

Firstly, AI detection systems have demonstrated their potential in numerous fields, such as fraud detection and health monitoring. So, there’s good reason to believe they could be valuable in an educational context as well.

Secondly, the fact that substantial investment is flowing into these technologies suggests that many believe in their potential. While investors aren’t infallible, it’s unlikely they would put significant money into a completely unproductive venture.

Thirdly, we don’t yet have comprehensive ‘gold standard’ research studies validating the efficacy of these AI detection programs in education. It’s all too new. What we do have, and only have, are smaller, preliminary studies. This situation indicates that the field is still emerging and evolving. Therefore, any scientific claims about these AI detectors, particularly those claiming they are useless, and especially those which don’t meet ‘gold standard’ research criteria, should be taken with a grain of salt.

Let’s also remember that detecting and interpreting complex human behavior is a challenging task, and expecting AI to be 100% accurate might be asking too much at this stage. But this doesn’t mean these tools are useless. Even if they’re partially effective, they can still provide valuable insights and aid in timely interventions.

Importantly, we should view these AI detectors as supportive tools, not as replacements for educators. They can help surface patterns that may be difficult for us to notice but still require our professional judgment to interpret and respond to.

Finally, we need to acknowledge the valid concerns around privacy and consent when implementing these tools. Ethical application and transparency are paramount in any technology use in education.

In conclusion, while AI detection systems have, at this point in time, been deemed “not perfect”, I don’t think it’s fair to dismiss them outright as ‘useless.’ They should be applied thoughtfully, ethically, and continuously improved upon through user feedback and ongoing research.

That said, since the launch of ChatGPT, I’ve not spend much time at all exploring the world of AI detection. Since November of 2022, I have single-mindedly channeled all my time and energy into exploring how my assignments need to change in the era of ChatGPT.