Executive Summary

New technologies have slowly become part of daily life. ChatGPT, a newly created chatbot, is no different. In the academic world, both educators and students are taking advantage of the tool and using it to their benefit. For educators, this means incorporating information they get from the chatbot into lessons. For students, this means using it as an additional learning resource. It can also mean using artificial intelligence to complete their work, raising concerns about academic integrity when using ChatGPT. This brief looks at the evolution of ChatGPT policies in schools, what not having any policies around it would mean, and evaluates policy options that allow students and educators to use the chatbot to their advantage and with integrity. 

‍

Overview

Whether it’s scrolling through social media or sending a work email, artificial intelligence (AI) is present in daily life. Most recently, the creation of ChatGPT has allowed users to generate responses to prompts that range from instructions on how to complete a task to writing an essay, raising questions on academic integrity in the education world. What is the best way for students and teachers to use ChatGPT? The most effective way allows students to take advantage of this new resource, but prevents them from using it in place of their own work. 

‍

Pointed Summary

• Students use ChatGPT for schoolwork assistance while educators are divided on how the chatbot should be used 
• Large K-12 school districts have banned the chatbot 

‍

Relevance

ChatGPT, created by OpenAI, was made available to users in November 2022. Since then, the software has been used to answer academic questions, draft emails and create essay responses. On their website, OpenAI writes that ChatGPT was created to interact in a “conversational” way. 

Additionally, in the time since its release, other tech companies (most notably Microsoft and Google) have rushed to put out their own interactive AI software.  

With the increase in access to softwares that generates responses, business leaders are becoming increasingly aware about the use of AI to respond to job application questions. Similarly, in the world of education, administrations, teachers and students are trying to find the most effective way to use this new, accessible artificial intelligence.   

While questions about how to use it are still being answered, the chatbot is currently being used by both educators and students. A Walton Family Foundation survey found that 22% of students use ChatGPT at least once a week and 40% of teachers use it once a week at a minimum. 1000 teachers and 1002 middle and high school students were polled in the survey. 

‍

Current Stances

Educators are divided on the use of ChatGPT in the classroom. One argument is that the AI chatbot is another technological evolution to teaching, similar to how YouTube and Google once were. 

Some teachers, who have used ChatGPT in their lessons, believe it is a teaching tool for educators to take advantage of. While reaping the educational benefits of the AI, these educators maintain that ChatGPT should be used solely as a source of aid, and should never be used in place of a student’s original work.

One negative effect educators see  is that because there is no well known way to determine whether or not a student’s response is actually from ChatGPT, students are free to use the chatbot instead of doing their own work, without punishment. 

The majority of students hold a similar belief, according to a new survey conducted by BestColleges. 51% of the 1000 undergraduate and graduate students polled said using AI created responses in their work falls under the category of cheating.  

‍

22% of students polled said they have used AI to help them with coursework or exams in the past and 32% said they intend to use or keep using AI for school assignments. 

With these contrasting views of students in the BestColleges poll, it can be inferred that while most students do not believe ChatGPT or other AI tools should be used, other students actively utilize them for their work. 

Moreover, in a poll of 216 students, 97% use ChatGPT or an AI tool in some capacity for their work. The level of AI employment, however, ranges from only using AI and not editing the answers provided to using the tool for minimal assistance on their work. 

‍

The use of AI tools for schoolwork is one of the reasons teachers oppose ChatGPT in academic settings. Students, however, are more inclined to favor the use of the tool for their work. The question that remains is where the line lies for cheating and not being academically honest in one’s work. 

When asked ‘How should ChatGPT be used in an educational setting?’ ChatGPT wrote that it should be used for assistance in work, including for research, writing conventions aid, translating aid, study materials, and tutoring. It believes that students should be able to take advantage of the tool to improve their learning, but “it's crucial to ensure that students understand how to use ChatGPT appropriately and ethically, and are not relying solely on it to complete their work.”

‍

‍

‍

The creator of ChatGPT, OpenAI, is also evaluating the effects the chatbot has on education and students’ learning processes. They are currently creating a new software that detects whether or not an answer was AI generated or written organically by a human. 

‍

Tried Policy

Different school districts are trying to find their ideal policy for ChatGPT usage. The most common policy, as of now, is to ban the use of ChatGPT entirely. 

Los Angeles Unified School District blocked ChatGPT in December to ensure their students used academic honesty in their work. 

Seattle Public Schools banned the chatbot on all devices owned by the district because they require original work done by students without any involvement of cheating. However, they plan to give ChatGPT access to educators in the district as a teaching resource.

Other districts that have banned ChatGPT include New York City Public Schools and smaller districts in Virginia and Alabama. 

Currently, no post-secondary institutions have created clear guidelines for how AI, including ChatGPT, can be used for coursework.

‍

Policy Problem

Stakeholders

Students are affected by policies about the use of ChatGPT in academic settings because they are in charge of presenting work that demonstrates their understanding of the content. Whether or not they have access to ChatGPT could affect the level of plagiarism they use and how much they understand the content, as the chatbot can be used as a learning tool. 

Educators play an important role in determining how to use ChatGPT because they set the rules for how students learn in their classes. Moreover, educators may choose to use the AI tool to assist them with teaching. This choice would affect how accessible the chatbot would have to be within a school district, which thus affects how accessible the tool is to students. 

Another potential stakeholder is the U.S. Department of Education. As the overseeing body for education in the United States, the department may decide to create policies around the use of ChatGPT in educational settings.

‍

Risks of Indifference

Since the beginning of the COVID-19 pandemic, students have been set back in their learning. Time that was meant to be spent on new content was spent reviewing old materials to ensure proper comprehension. This is one of the reasons why a learning gap exists for K-12 students. 

Used as a learning tool, ChatGPT has the potential to decrease the gap created by the pandemic. However, if it is used to do the work that would help students’ learning instead of a tool of aid, the learning gap will grow instead of shrink. 

An increased learning gap, paired with the effects of the pandemic could be detrimental to the American economy in the future. 

The majority of American schools returned to complete in-classroom learning at the beginning of the 2021-2022 school year. According to an analysis done by McKinsey & Company, this could result in a $483 billion loss to the United States’ gross domestic product (GDP), which measures a country’s economic output. 

Without proper regulation of ChatGPT employment in educational settings, students could take advantage of the chatbot and generate responses to questions instead of writing their own. The failure to complete tasks themselves gives students greater learning loss, which hurts how they work in the future. As the next generation to enter the workforce, missing skills will appear in the work conducted and affect the entire American economy negatively. 

Therefore, it is imperative that the utilization of ChatGPT in schools is monitored so students can benefit from using the chatbot as a teaching tool, but are limited in how much of their work can come from the chatbot. 

‍

‍

Nonpartisan Reasoning

The goal, above anything in politics, is to keep the country going and growing. Without the proper monitoring of ChatGPT usage in academic settings, students may not learn as much as they would have without the chatbot that has all the answers.

This learning gap, added to the learning loss of the COVID-19 pandemic, will hurt the country’s economy in the future (see Figure 4). 

Thus, instead of leaving the economy to face greater damage than it may already, a policy tracking ChatGPT use in academic settings should be put in place. This way, educators and students can utilize the chatbot for aid, but not use it to do all their work. 

‍

Policy Options

One route schools could go is what many are already doing: block ChatGPT altogether. This limits the access students and educators have to the chatbot and will lower the risk of plagiarism at school. Still, students will have access to ChatGPT on personal devices and may still use it to complete homework. 

The school districts who have banned ChatGPT are trying to find ways to make it accessible to educators as a teaching tool. Giving educators access to the chatbot will help them find new ways to teach and match different learning styles. Here though, if educators are allowed to use ChatGPT, it cannot be completely blocked from school networks, thus giving students the opportunity to be able to use it. The benefits and costs of that must be weighed by individual districts. 

The third option is to focus on the benefits of ChatGPT and make it accessible to all students and educators. Educators will be able to use the chatbot as a teaching tool and be able to explain concepts in new ways. Students will be able to use it to teach themselves difficult concepts and fix their grammar, among other things. 

While plagiarism is always an option, there is artificial intelligence being created to determine whether or not a text is chatbot generated. OpenAI is developing one software, which, coming from the same company as ChatGPT, will most likely have strong detection tools. Another detection tool, called GPTZero, was created by Edward Tian, a computer science student at Princeton University. GPTZero, since its release to the public, has not been perfect at detecting AI generated texts. Nonetheless, teachers have been using it and according to Tian, have told him about their positive results. Tian is actively working to make the software more accurate, which only increases the ability to check if a student’s plagiarism is generated by a chatbot like ChatGPT.

Conclusions and Recommendations

There are still several unknowns about how to use ChatGPT. The most effective way to use the chatbot is as a resource for students and educators, following the third policy option. Not only do educators get an additional teaching tool, but students get a learning tool that can match their learning style and needs. Moreover, despite the concern over plagiarism, there are softwares, available to the public, built to detect AI generated texts. With this option, the main players in the academic world get to take advantage of ChatGPT, a new and developing tool. 

‍

‍

Acknowledgement

The Institute for Youth in Policy wishes to acknowledge Gwen Singer, Sarah Zhang, Paul Kramer, Carlos Bindert and other contributors for developing and maintaining the Effective Discourse Department and associated Fellowship programming.

References

1. Bowman, Emma. 2023. “A College Student Created an App That Can Tell Whether AI Wrote an Essay.” NPR, January 9, 2023, sec. Technology. https://www.npr.org/2023/01/09/1147549845/gptzero-ai-chatgpt-edward-tian-plagiarism.
2. Dorn, Emma, Bryan Hancock, Jimmy Sarakatsannis, and Ellen Viruleg. 2020. “Achievement Gap and Coronavirus | McKinsey.” Www.mckinsey.com. June 1, 2020. https://www.mckinsey.com/industries/education/our-insights/covid-19-and-student-learning-in-the-united-states-the-hurt-could-last-a-lifetime.
3. Gecker, Jocelyn. 2023. “Some Educators Embrace ChatGPT as a New Teaching Tool.” PBS NewsHour. February 14, 2023. https://www.pbs.org/newshour/education/some-educators-embrace-chatgpt-as-a-new-teaching-tool.
4. Goldberg, Suzanne. 2021. “Education in a Pandemic: The Disparate Impacts of COVID-19 on America’s Students.” https://www2.ed.gov/about/offices/list/ocr/docs/20210608-impacts-of-covid19.pdf.
5. Hao, Karen. 2023. “What Is ChatGPT? What to Know about the AI Chatbot That Will Power Microsoft Bing.” The Wall Street Journal. March 23, 2023. https://www.wsj.com/articles/chatgpt-ai-chatbot-app-explained-11675865177.
6. Jimenez, Kayla. 2023. “As Battle Persists over AI, Here’s What Teachers, Students Have to Say about ChatGPT.” USA TODAY. March 1, 2023. https://www.usatoday.com/story/news/education/2023/03/01/what-teachers-students-saying-ai-chatgpt-use-classrooms/11340040002/.
7. Johnson, Arianna. 2023. “ChatGPT in Schools: Here’s Where It’s Banned—and How It Could Potentially Help Students.” Forbes. January 18, 2023. https://www.forbes.com/sites/ariannajohnson/2023/01/18/chatgpt-in-schools-heres-where-its-banned-and-how-it-could-potentially-help-students/?sh=62e0b1e06e2c.
8. Kakade, Angeli. 2023. “Here’s Why ChatGPT Is Blocked on All Seattle Public Schools Devices.” King5.com. 2023. https://www.king5.com/article/news/education/seattle-public-schools-block-ai-powered-chat-gpt-school-devices/281-d8a2e201-b215-4786-8b2d-77487b8d6397.
9. Marr, Bernard. 2019. “The 10 Best Examples of How AI Is Already Used in Our Everyday Life.” Forbes. December 16, 2019. https://www.forbes.com/sites/bernardmarr/2019/12/16/the-10-best-examples-of-how-ai-is-already-used-in-our-everyday-life/?sh=15365ed1171f.
10. OpenAI. 2022. “ChatGPT: Optimizing Language Models for Dialogue.” OpenAI. OpenAI. November 30, 2022. https://openai.com/blog/chatgpt.
11. Rosenzweig-Ziff, Dan. 2023. “New York City Blocks Use of the ChatGPT Bot in Its Schools.” Washington Post, January 5, 2023. https://www.washingtonpost.com/education/2023/01/05/nyc-schools-ban-chatgpt/.
12. Welding, Lyss. 2023. “Half of College Students Say Using AI Is Cheating | BestColleges.” Www.bestcolleges.com. March 27, 2023. https://www.bestcolleges.com/research/college-students-ai-tools-survey/.
13. Yang, Maya. 2023. “New York City Schools Ban AI Chatbot That Writes Essays and Answers Prompts.” The Guardian, January 6, 2023, sec. US news. https://www.theguardian.com/us-news/2023/jan/06/new-york-city-schools-ban-ai-chatbot-chatgpt.

‍