By Mariano Felice, British Council
BritishCouncil_pole

 

The British Council published a position paper outlining ten principles for a more responsible, fairer and inclusive application of AI in English teaching, learning and assessment.

AI needs to be understood as an enabler rather than a driver. Research, AI literacy, and broader, more diverse access to these tools, are also important.
The British Council, the UK’s organisation for cultural relations and educational opportunities, has unveiled its recommendations for a more responsible, fairer, and inclusive application of Artificial Intelligence in education – specifically in English learning, teaching and assessment. The paper, Human-centred AI: Lessons for English Learning and Assessment, delves into the need of a more human-focused approach to the selection and implementation of AI technologies. This approach is rooted in research and educational purpose, with the needs of learners and greater inclusivity at its core.

The role of teachers, collaboration between experts from various fields, and the AI literacy of all stakeholders within the education system, are all fundamental for applying AI in ways that enrich and make the learning and assessment experience more flexible. Ensuring access to these new technologies to democratise their use and leveraging their full potential to personalise experiences for learners with diverse education needs are also key areas explored in the British Council’s work.

On the importance of adopting a more human, professional, and rational perspective towards AI in education, Dr Mariano Felice, author of the paper and senior researcher and data scientist for Language Assessment & Learning English at the British Council, says: “We cannot leave crucial decisions, such as education, in the hands of a machine, as there are regulatory and legal matters to address. Nor can we separate morality and ethics from the field of education. Technology should be used as a tool to assist us, but experts must always oversee its use to prevent and correct any deviations.”

The ten principles for a human-centred approach to English learning and assessment

  1. Embrace the possibilities. AI is a transformative technology. In the context of language teaching, learning and assessment, AI can push the boundaries of what has been possible until today, enabling new approaches that can make the whole learning journey more engaging, personalised and efficient. Organisations that fail to move with the times and harness this potential will find it increasingly difficult to remain relevant.
  2. Remain purpose focussed. As key players in the language teaching, learning and assessment landscape, we must resist the idea of using technology without a clear purpose or strategy. The integration of AI into teaching and assessment practices can be greatly beneficial, but only when it’s built on research and the learning goals have been prioritised.
  3. The ecosystem is key. Acknowledge that curriculum, delivery, and assessment must work together in a Comprehensive Learning System. To ensure that technology becomes a facilitator and not a barrier, it must integrate seamlessly into the whole education ecosystem. This, as all systems, operates within a context of use defined by its key stakeholders, with the learner recognised as the most important.
  4. AI literacy for everyone. It is essential that all stakeholders learn fundamental concepts in AI, how it works, and how it is used in teaching, learning and assessment products and services. The level of literacy required by different stakeholders will depend on what they need to know to make informed decisions in their role within the learning system and should empower them to critically assess tools before their adoption.
  5. Humans before technology. The British Council’s approach to AI is human-centred, prioritising the needs and well-being of learners before all else. All technology, including AI, should only be applied to the learning and assessment system if it can be proved to enhance the experience and outcomes for the learner or test-taker.
  6. Educators are essential. The intelligent capabilities of Generative AI, and more precisely large language models such as ChatGPT, cannot sweep away the body of knowledge and experience that educators and language experts have built over centuries. By adopting a balanced approach where we combine the best of AI with human expertise, we can make sure we deliver modern and efficient solutions that enhance the educational experience for all stakeholders.
  7. Collaborate with experts. In an age of constant technological change, it is challenging for non-experts to keep up with the latest developments. The British Council recognises the importance of different disciplines in shaping the future of education, and actively encourages interaction with experts inside and outside our field of expertise.
  8. Commit to ethical standards. Ethical considerations should be a priority for providers and users of AI systems, as they encourage us to use technology in a way that is fair, safe and inclusive. The British Council’s ‘ethics by design’ approach aims to identify and address potential ethical concerns from the early stages of development.
  9. Ensure solutions are truly accessible. Education systems need to ensure that AI applications are available on devices and in forms which are widely accessible. The design and application of AI should not be predicated only on the use of those who already occupy privileged positions in society but should be created to actively engage learners irrespective of their background.
  10. Be inclusive and embrace diversity. It is essential that AI models are built using balanced and diverse datasets to ensure that different groups are represented without bias or discrimination. We must also ensure that AI is used to increase opportunity for the full diversity of international learners, particularly those with cognitive and physical disabilities. AI has the capacity to support learners by personalising the learning and assessment journey to their needs.

Human-centred AI: lessons for English learning and assessment:
https://www.britishcouncil.org/research-insight/human-centred-ai-lessons-english-learning-assessment.