As schools worldwide wrestle with how to respond to artificial intelligence, from outright bans to cautious experimentation, teachers at the International School of Kenya decided to do something different: teach with it, not against it. Rapid expansion of generative AI is leading schools to fret about students’ cheating. Some educators are adapting by reconfiguring once standard assignments like book reports. Others have openly talked about banning generative AI. A group of social science teachers at the International School of Kenya (ISK) opted for a different approach. Three ISK teachers revamped an introductory Grade 9 social science unit in order to equip students to analyze the usefulness and accuracy of generative AI platforms.
“Teachers should explicitly teach learners about AI use in social science, including its dangers and opportunities,” according to ISK teacher Sekou Otondi. He along with colleagues Rachel Hopkins and Samuel J. Richards brainstormed to create opportunities for students to respond wisely to rapidly expanding AI technology. They collaboratively designed “Discovering Nairobi and AI” as a unit using two inquiry cycles for students to test their hypotheses. The first inquiry cycle focused on their host city of Nairobi, while the second cycle allowed students to choose another city that they knew well.
“The goal was to test AI, not to research using AI,” according to Richards. “This was an important distinction to reinforce with our students. It also took them a while to understand this important difference.”
Students first learned about the origins and development of Nairobi using more traditional sources. This included maps, census data, historic photographs, and excerpts from written histories of Kenya’s capital city, such as Bethwell and Madara Ogot’s 2020 text History of Nairobi: from a railway camp and supply depot to a world-class African metropolis. Once students gained foundational knowledge about the origins and development of Nairobi they were ready for the next step.
Students were then introduced to relevant concepts like Large Language Models (LLMs), engineering prompts, hallucinations, and blind spots. They also read an excerpt from Kelsey Rice’s “Chat GPT and World History Essays” published in Teaching History: A Journal of Methods (winter 2024). For Richards, “Rice’s article breaks through the advertising noise of AI in the media to provide the focused critique of a historian. Her work is evidence-based but had to be adapted for Grade 9. It is a perfect provocation for an inquiry cycle.
An excerpt of Kelsey Rice’s analysis of ChatGPT and bias included in the Social Science 9 reader for students at the International School of Kenya in Nairobi. (Photo source: Kelsey Rice)
Rice investigated the use of ChatGPT with university-aged students at Berry College in Georgia, USA. She argues that generative AI has patterns of bias borne from LLMs data sets which tend to amplify Eurocentric, misogynistic, and white supremacist sources. That’s a troubling finding, especially for readers in Africa.
“Rice used big vocabulary words and heavy ideas for Grade 9. Understanding these potential flaws in data sets also invited students to scrutinize results more closely,” according to Richards.
Students were equipped to analyse answers to their engineered prompts with heightened awareness for distortions.
“I feel the use of AI in the Nairobi inquiry unit has opened students’ eyes beyond the 'innocence' of AI as just another quick go to tool for research. It exposed them to the power dynamics in favor of and in disfavor of certain global regions,” according to Otondi.
During both inquiry cycles, students used the same engineered prompt for two different generative AI platforms. Popular platforms that students chose included Perplexity, ChatGPT, Grok, Deep Seek, Gemini, and You.
Hopkins points out that, “The learning worked best when it was broken into steps, since AI can feel overwhelming if it’s too open-ended. Scaffolding really helped students understand the process more.” Teachers relied on graphic organizers to help students through the two inquiry cycles: first with Nairobi, then with a city of their choosing.
Hopkins saw this repetition as crucial. She reflects that “This aspect was key. It helped with the transfer of skills by allowing students to work through the same process but in a different context.”
Teachers planning similar units might learn from the successes and challenges of the Grade 9 work at ISK.
“Students ran into two main challenges when working with AI,” according to Hopkins. “The first was figuring out how to tell the difference between actual facts and the hallucinations AI can give. The second was realizing how much the quality of the answers depended on how they worded their prompts, which was both a good learning moment and a bit frustrating. What was really interesting, though, was that many of them decided on their own to go back and reword their prompts just to see how the answers would change. In doing so, they started pushing back against the AI, almost like they were trying to catch it in a lie, which really showed how engaged and critical they were becoming.”
As a final task, students responded to the prompt:
Based on your experience with two AI platforms, recommend to the ISK Social Science faculty the appropriate student usage of AI and LLMs. Be sure to justify your recommendation with evidence from your investigation of ____ [city].
ISK students responded with a range of recommendations; they had mixed views much like educators around the world. For instance, one student who analyzed the quality of answers from ChatGPT and Grok regarding the origins of the Malaysian capital city of Kuala Lumpur found that both AI platforms provided “in-depth answers about the indigenous and colonials” who contributed to the growth of Kuala Lumpur. He argued these answers were “...accurate and refuted many claims that Gen AI has a bias and is blind to indigenous perspectives.” He suggested that “The Social Science Department at ISK should allow and even teach students to use generative AI and LLMs.”
However, another Grade 9 student investigating the Paraguayan capital of Asunción cautioned that “The ISK social science faculty should heavily caution students when using AI platforms for research due to its bias towards all things European and its hallucinations.” Her inquiry cycle using ChatGPT and Perplexity led her to conclude that those platforms “responded with mainly Spanish colonialists' perspectives and not the [indigenous] people of Paraguay.” She found that missing perspectives of the Guarani people represented a significant AI blind spot.
Whatever policies your school adopts regarding the use of generative AI, Otondi says, “Teachers should keep in mind that AI generative tools are here to stay and might be much better and more advanced as time goes on. It is important to ensure the ethics of research, even more so in social science. However, primary sources will always provide a ‘lived experience’, something generative AI will never be able to provide as part of authenticated research.”
The work at ISK reminds us that the question isn’t whether AI belongs in the classroom but how we can teach with curiosity, ethics, and discernment in a digital age.
Samuel J. Richards is an experienced International Baccalaureate (IB) history and Global Politics teacher currently based in Nairobi, where he serves as a high school social sciences teacher and head of department at the International School of Kenya. He enjoys students’ “lightbulb moments” and has a passion for literacy learning, the humanities, and curricular improvement. He holds a master’s degree in history and Master of Science in Education in curriculum and instruction. His career spans public and private schools on five continents, teaching Grades 6-12 including Advanced Placement and IB courses, undergraduate history, and mentoring pre-service teachers.
Sekou Otondi teaches IB Global Politics and Social Sciences at the International School of Kenya. He is a researcher and educator based in Nairobi, specializing in global politics with a particular emphasis on regional peace and security in Africa. He holds a Master of Arts in international studies from the Institute of Diplomacy and International Studies at the University of Nairobi, where he is currently pursuing a Doctor of Philosophy in political science and public administration. In 2016, he won the Tana Forum Essay Competition. In addition, Sekou’s insights have been featured in prominent publications such as The Conversation Africa, African Arguments, and Africa is A Country. He has also contributed to key discussions at notable conferences, including the fifth Tana High Level Forum on Security in Africa in 2016.
Rachel Hopkins is an international educator and school leader currently serving as interim Assistant Principal of the high school at the International School of Kenya. She has taught IB Psychology and other social science courses in both public and private schools. She is especially passionate about student wellbeing, academic growth, and inclusive learning communities. Throughout her career, she has prioritized student leadership and holistic education. She holds a Master of Arts in Teaching and is trained in multiple IB subject areas, AP Psychology, and technology-enhanced learning tools. She is currently pursuing a Doctor of Education in international educational leadership.
