Governments Urged To Get A Grip On AI In Schools

Governments have been urged to get a grip on AI in schools to help protect students and teachers.

And one recommendation in new guidance launched today is the introduction of an age limit of 13 on the use of generative AI in education.

“Generative AI can be a tremendous opportunity for human development, but it can also cause harm and prejudice,” said Audrey Azoulay, Director-General of UNESCO, the UN’s education and cultural arm which produced the new guidance.

“It cannot be integrated into education without public engagement, and the necessary safeguards and regulations from governments.”

The prospects of AI having a far-reaching impact on education has rocketed up the policy priority list ever since Open AI demonstrated its potential with the release of ChatGPT in December last year.

Concerns have revolved around the risk of students using AI to cheat in coursework, as well as the effect on children’s mental health.

But these have been counterbalanced by hopes that AI could have a positive impact in schools, including teaching critical thinking and editing work.

Now the UNESCO guidance aims to set out the steps governments should take to regulate the use of AI and set out the ground rules for its ethical use in education and research.

A major concern is that existing AI models are trained using data from online users, which risks worsening digital divides and over-emphasizing the values of the Global North.

The rapid pace of development of AI has caught many education leaders off-guard.

A UNESCO survey earlier this year found that just one in 10 schools and universities had formal guidance on the use of AI, and the organization warned in June there was a worrying lack of public scrutiny over AI in education.

Key elements of AI regulation include creating a national body to co-ordinate the government’s approach to AI, and ensuring regulations are aligned with existing legislation covering areas such as data protection, privacy and internet security.

Governments are also urged to seek to find a balance between regulation and promoting innovation, including promoting sharing resources and creating high-quality content for the public good.

The guidance warns that AI applications “often entail substantial risks for children, including exposure to inappropriate content as well as the potential for manipulation” and recommends an age restriction on their use to protect children’s health and well-being.

While UNESCO puts the recommended age at 13, the guidance also notes that some experts argue that this should be raised to 16, and acknowledges there are concerns over whether self-reporting age is appropriate or whether additional verification needs to be introduced.

“As GenAI tools are increasingly able to automate some basic levels of writing and artwork creation, revisiting why, what and how we learn are now critical considerations for education in this new phase of the digital era,” said Stefania Giannini, UNESCO’S assistant secretary general for education.

“It is my hope that this guidance will help us redefine new horizons for education and to inform our collective thinking and collaborative actions that can lead to human-centred digital learning futures for all.”

The guidance is released during UNESCO’s Digital Learning Week, which includes talks by speakers including Yoshua Bengio and Yann LeCun, two of the triumvirate often described as the “godfathers” of AI.

While Professor Bengio has previously said he regrets not giving safety a higher priority during the development of AI, Professor LeCun, chief AI scientist at Meta, has described apocalyptic warnings over the potential impact of AI as overblown.