Bridge or barrier. Does generative AI contribute to a more culturally inclusive higher education and research?

Indicative image shows people crossing a golden bridge supported by a large concrete hand.

While the ability of generative AI to produce text in English has been widely reported, the implications of its ability to translate and act as a cultural mediator into English have received less attention. Given the contexts of use in education and research, Dmitrinka Atanasova, invites higher education policymakers to use a cultural lens when developing policy responses to generative AI.

Since the release of ChatGPT, there has been much discussion about the value of AI-enabled writing tools for higher education and scientific publishing. declare student essay deadand the surge of its use for writing Journal article. Academics continued to express various, sometimes conflicting answersfrom direct prohibitionsto demonstrate how ChatGPT can help create reliable estimates. Because AI tools need to be integrated into Microsoft Office and key learning platforms such as Moodle, generative AI will be even harder to ignore. However, in recent discussions, one aspect has not been given due attention – its connection with cultures and the ability to build bridges between them.

ChatGPT and culture

This does not mean that the connection between ChatGPT and culture is ignored. Academics began to call for more research explore cultural biases in responses generated by ChatGPT. Inevitably, this type of research are starting to appear. The researchers also recommended the use more linguistically and culturally diverse training data as part of the continuous improvement of ChatGPT.

Recent debates about the value of ChatGPT for higher education have also included views that a key benefit of this tool could be help in learning a foreign languageoften described as Gateway to intercultural competence or the ability to understand and respect each other despite language and cultural barriers. Early research also shows that ChatGPT is already a good translator. different languageswhich promises intercultural communication.

From the very early days of the release of ChatGPT-4, it has also become clear that not all cultures have the same views on AI-enabled writing tools. Italy recently banned ChatGPT along with China, North Korea, Iran and Russia. In other countries, such as the UK, the government has taken a more liberal stance and instead called on regulators to “come up with individual, context-sensitive approaches‘.

These divergent responses to ChatGPT are not unexpected. Project Global AI Narrativesfor example, shows how perceptions of the risks and benefits of AI are shaped by cultural norms and values.

ChatGPT and culturally inclusive higher education?

Different attitudes towards ChatGPT in different countries – bans or openness – may have implications for higher education in the UK and other countries that have a more open national policy towards ChatGPT. Different attitudes can exacerbate disparities between local and international students. While UK students may have had the opportunity to work with ChatGPT, international students from countries where such tools are banned have no prior experience.

Different attitudes can exacerbate disparities between local and international students

As with any new technology, this highlights the importance of providing sufficient support to help learners from all walks of life make good use of such tools. When Turnitin’s plagiarism detection service was launched, universities debated whether or not to make it available for students to use. As universities began to appreciate the educational potential of enabling students to use Turnitin, the focus shifted to offering students help in interpretation their turnitin records.

ChatGPT can also be useful for international students who often report having difficulty expressing themselves in a foreign language such as English, but at the same time clarity of expression, grammar, and style are assessed by the writing assignment components. Students could copy their entries into ChatGPT with a hint to improve the grammar and style of the posted text. Thus, ChatGPT can function as a useful text editing tool. However, it also suggests a need to reassess how these aspects of academic writing replace elements of cultural learning in assessments.

ChatGPT and culturally inclusive academic publishing?

ChatGPT can also help researchers polish their manuscripts before submitting them to journals. While biodiversity is essential to preserving research on a wide range of topics in vernacular languages, and there is some evidence that an open and bibliographic publishing ecosystem already exists, English continues to dominate the top-ranking academic journals.

Academic researchers early in their careers, in particular, report significant pressure to publish in prestigious English-language journals, but find it particularly difficult to do so. This potentially creates a barrier for the next generation of researchers with lower levels of English proficiency. Generative AI can help lower this barrier by acting as a text editing tool.

As a powerful cultural mediator in English, it lowers barriers to entry but at the same time redefines how we understand intercultural learning.

However, this potential use of ChatGPT raises difficult questions about the protection and promotion of biodiversity. By allowing researchers to adapt to the English language publishing system, tools such as ChatGPT could help reinforce and accelerate the trends that position English as lingua franca of global research and place English at the top of the existing hierarchy of scientific publications.

ChatGPT can be a useful tool for international students and researchers trying to write eloquently in English. But its implications for cultural inclusion are more ambiguous. As a powerful cultural mediator in English, it lowers barriers to entry but at the same time redefines how we understand intercultural learning. For higher education policy makers, this points to the need to be culturally sensitive when designing responses to generative AI.

The content created on this blog is for informational purposes only. This article represents the views and opinions of the authors, but does not reflect the views and opinions of the Impact of Social Science blog (blog) or the London School of Economics and Political Science. Please see our comment policy if you have any concerns about posting a comment below.

Image credit: to nag via

Printable, PDF and Email