Initially, when AI was released, the reaction within education was to restrict all use, which is understandable. After all, no one truly understood its capabilities, reliability, or how to define what utilizing AI meant within the context of student assignments. Was it cheating? Was it a resource tool? No one knew.
But now, it’s been over a year since the release of OpenAI’s ChatGPT, and more AI models have arrived on the market and become part of our daily lives. In November of 2023, OpenAI announced the expansion of its product offerings, including customizable ChatGPT bots. Google has fully integrated its AI, Bard, into their search engine, which has nearly 62% of the browser market share.
Bottom line: It’s time we learn how to work with, not against, AI.
Never is that more prevalent than within education.
That’s why we’ve decided to answer a few key questions we often get in an effort to help open the dialogue between educators and students about utilizing AI and AI content detectors.
How should I approach the topic of AI within my classroom?
There are a few essential points when approaching the topic of AI.
First, it’s important that before AI is brought into the classroom, guidelines and rules are put in place so that all faculty and educators are on the same page. Reactions to AI echo that of the calculator. At first, it was banned outright, but over time, certain rules and guidelines were put in place about when a calculator could be used in the classroom and when it couldn’t. Look at addressing AI in a similar manner.
Second, don’t deny that AI exists. AI is here and isn’t going anywhere. Therefore, be realistic about AI and accept that it might be time to figure out how to optimize its potential. In fact, educator professional development should consider including generative AI training so teachers and faculty are familiar with its possibilities and shortcomings.
If you want to embrace AI in your classroom, do it with your students. Take time to use AI models in the classroom. Input questions and prompts and go over the responses with students, highlighting where there could be inaccuracies or where students would need to do more research that isn’t AI and show their work of doing so.
Furthermore, reviewing questions and prompts within the classroom can help students learn how to utilize AI for research properly and not to believe everything it generates as fact.
In short, bringing AI into the educational fold should be a joint effort between educators and students.
What should I do if AI content is detected within a student assignment when the guidelines state no AI-generated content can be used?
As a result of analyzing millions of documents over the years, the Copyleaks AI Content Detector has an overall accuracy of over 99% and the lowest false-positive rate of any platform, at 0.2%. Nevertheless, they can still happen despite the best efforts to minimize false positives.
We encourage educators to use the Similarity Report as a data point to discuss how a student might have worked on the assignment. It’s important to investigate further into the assignment and ask critical questions, such as:
- Does this style match that of previous assignments from the student?
- Is the writing style inconsistent?
- Is there a robotic feel to the assignment instead of a natural flow?
- Is the writing/vocabulary above the student’s grade level?
It’s crucial to remember that AI detectors help bring attention to a potential concern. It’s also important to note that the Copyleaks AI Content Detector differs from other detectors in that each scan seeks to verify human-written, not AI-generated. So when an alert for AI comes up on a Similarity Report, it should be seen as the system being unable to verify that a human wrote the content scanned; it is not declaring the AI content is 100% present.
Therefore, it should always be left to the educator to investigate further and open a dialogue about AI usage with a student to help assess whether or not it falls within the guidelines about utilizing AI for the assignment.
How can I ask my students to prove they didn’t use AI for an entire assignment?
When it comes to having the conversation around AI and whether or not a student uses it to write their entire assignment or only uses it as a resource, it helps to apply some old-school methods of “show your work.”
For instance, if an assignment is flagged as not being fully human-written, you can meet with the student one-on-one and ask them to discuss the answers or write a synopsis of their paper on the spot. Did they touch on all the points covered, or are they uncertain? If it’s the latter, chances are higher that they didn’t entirely write the assignment.
You can also keep track of each instance of potential AI detection and determine if a student shows a track record of AI use. If so, you can discuss concerns with the student and include the historical data to make the conversation productive.
Another option is if they are allowed to use AI as a resource, to ask them to supply the prompt that was used within the AI model for research and discuss what questions they asked the AI, then take it a step further and go over their process in doing their research beyond AI.
AI Is Constantly Evolving, So Are AI Detectors
As AI models continue to evolve, so do AI content detectors. Regarding the Copyleaks AI Content Detector specifically, we never see it as a finished product.
Walt Disney once said that Disneyland would never be completed but would continue to evolve and grow. We feel the same way about the AI Content Detector, such is the nature of AI in general and the conversations around it. Our algorithms continue to learn as AI technology evolves, increasing our already accuracy rate even more.
Furthermore, we’re always listening to our educators, and as a result, Copyleaks will soon release an analytics feature as part of our LMS integrations to provide educators with a more detailed view of scan results and supply historical data around if there is a pattern of plagiarism or AI use. With that insight, a teacher will have all the necessary information to make the best decision before opening up the discussion with a student.
So, as conversations begin to occur around bringing AI into the classroom, it’s crucial to remember that it should always be about moving forward, with the flexibility to adapt as the technology evolves.
But most of all, it is essential to view the process of utilizing AI in education as a collaboration between educators and students. Go on the journey together with open communication.
Because the reality is we’re all still learning about AI as we go. So, let’s go together.