The Impact of AI on History Education: A Critical Examination of ChatGPT’s Eurocentric Bias

Title: The Impact of AI on History Education: A Critical Look at ChatGPT’s Western-Centric Bias

In the realm of education, the integration of artificial intelligence (AI) has been touted as a game-changer, promising to revolutionize the way students learn and teachers instruct. However, a recent study has shed light on the potential pitfalls of relying on AI in history education, particularly when it comes to perpetuating a Western-centric bias.

At Harvard University, for example, the introductory computer science course incorporates AI-based platforms to guide students in learning programming. Additionally, a university pedagogy project aims to help educators critically engage with AI in teaching, while workshops on harnessing the power of AI are offered to faculties and teaching fellows.

As history educators, the idea of integrating AI into history classes seemed promising. We decided to put ChatGPT, a popular AI language model, to the test by asking it to provide an outline of world history. While ChatGPT was able to generate a comprehensive outline divided by historical periods, starting from prehistory to the contemporary era, we noticed a glaring omission in its coverage of non-European history.

When ChatGPT delved into the classical and early modern periods, the focus on non-European history gradually diminished, with the outline predominantly centered on European events. Despite our attempts to prompt ChatGPT to include more diverse historical narratives, it consistently defaulted to a Eurocentric perspective.

A deeper investigation revealed that ChatGPT’s training on an English-language dataset has led to a Eurocentric reading of world history. Its understanding of historical events is largely Western-centric, with a notable lack of representation of Asian and non-Western histories.

This Western-centric bias in educational technology raises concerns about the impact on students’ worldview and critical thinking skills. By defaulting to a single Anglo-American value system, ChatGPT risks universalizing Western knowledge at the expense of diverse perspectives and narratives.

If AI technologies like ChatGPT become the default mode of education for history, there is a real risk of perpetuating indifference towards Asian and non-Western history. Efforts to promote diversity, equality, and global history could be undermined, leading to a moral indifference to global inequality and historical injustices.

While AI has the potential to democratize education and provide greater access to knowledge, it is crucial to address the inherent biases and limitations in current AI models. A collaborative effort between technologists and humanities scholars is essential to develop responsible AI for education that reflects a more inclusive and diverse understanding of history.

In conclusion, the rise of AI in education marks a critical juncture for educators to rethink teaching and testing methods. By acknowledging and addressing the Western-centric bias in AI technologies, we can strive towards a more equitable and inclusive approach to history education.

Queenie Luo, a PhD candidate at Harvard University, and Michael Puett, a professor of Chinese history and anthropology at Harvard, are leading voices in the discussion on the impact of AI on history education. Their insights shed light on the importance of critically examining the role of AI in shaping students’ understanding of the world.