Response to the Government Call for Evidence on Generative AI in Education

The Alan Turing Institute welcomes the Department for Education’s statement on generative artificial intelligence in education, as we believe this is a critically important area that deserves significant attention.

The response synthesises the perspectives of researchers at the Turing with expertise and interest in the area of artificial intelligence (AI) within education, and in ethical and responsible innovation and governance. The Turing’s Public Policy Programme drafted the response and compiled views from the Institute’s wider academic community. 

Our Call for Evidence discusses the Government's questions in depth, including these main points:

• The concerns and risks regarding the use of generative AI in education include: dependency on commercial generative AI, infringement of rights, the spread of bias, misinformation, and harmful content, plagiarism, and social harms such as the increased passive consumption of information, overreliance on technology, and the widening of the digital divide.
   
• There are multiple ways in which generative AI could benefit the education sector, including offering assistance on administrative tasks, bolstering language education, providing support for neurodivergent students, and improving children’s data literacy and critical understanding of AI technologies.
   
• There is a need for a widespread understanding that generative AI is not a one-size-fits-all solution to issues in the education sector, and the use of the technology should not replace creativity, originality, critical thinking, and collaboration amongst students and teachers.
   
• We call for an in-depth analysis of off the shelf LLMs, as research has indicated that many of the available offerings are not suitable for use in the classroom, often due to their lack of proficiency in school subjects.
   
• Children and young people must be meaningfully engaged on the topic of generative AI. We highlight an example of this work from The Alan Turing Institute, Exploring Children’s Rights and AI in collaboration with the Children’s Parliament and the Scottish AI Alliance, which has been engaging children and young people across Scotland and serves as an example of stakeholder co-design and engagement.
   
• We believe there are many ways in which the DfE could help to support students, teachers, parents, and guardians. This includes providing clear guidance on the advantages and disadvantages of using generative AI in an educational setting, providing trainings for education staff, applying a vetting process to the use of generative AI in classrooms similar to current curricula procedures, carrying out meaningful engagements with relevant stakeholders, and ensuring safeguards are in place to protect children and young people.

We look forward to learning about the DfE's next steps with regards to generative AI in education.