The Cambridge approach to generative AI and assessment

Cambridge has always used technology to empower educators, and AI will be no different.  Our approach and policies on AI will evolve as the technology develops and as Cambridge and our partners, especially schools, colleges and teachers, establish best practice.

OCR Chief Executive Jill Duffy said: “Quality and integrity will come first, whatever technology we use. We will listen to schools and colleges and support them in developing best practice for teaching and learning.”

 
Rod Smith, Group Managing Director for International Education at Cambridge, said: “Generative AI offers opportunities and challenges for teaching, learning and assessment - and it's essential that we prepare now. 

“We are developing key principles and practices to govern robust and ethical deployment of AI. Schools, teachers and learners face practical questions that can impact on access, equity and integrity. 

“We will work with our 10,000-strong community of schools worldwide as they engage with this rapidly evolving technology. That will mean embedding the knowledge, skills and conceptual understandings of AI into an adaptable and cutting-edge curriculum that supports a fairer future.”

Strengthening professionals, supporting teachers

Generative AI and large language models, like ChatGPT and Bard, can enable Cambridge, our partners, researchers, educators and students to achieve more. 
 

• We will augment the role of people and uphold the value of established methods. It’s education professionals who will enable us to take the fullest advantage of the AI opportunity. Used well, AI can strengthen their irreplaceable expertise for the benefit of students. 
• AI and large language models can be used as effective teaching tools in classrooms in appropriate circumstances. Cambridge will support and listen to schools and teachers as they carefully and critically engage with such resources and tools, while upholding standards. 
• Students can develop critical skills in the classroom to equip them to flourish in a world of big data, AI and large language models, including distinguishing between appropriate and inappropriate use of AI. This builds on established evidence, teaching principles and methods in combination with rapidly developing technologies.

Delivering better assessment 

• In many cases, the rise of AI reinforces the value of in-person exams, as well as digital exams with human supervision, which are fair, rigorous and help guard against plagiarism and cheating.  
• As with other forms of plagiarism and malpractice, students must not pass off AI generated work as their own.  
• Thoughtful use of technology improves how we develop and deliver assessment. Our pioneering use of digital exams shows how we are doing just that.
• We can take the best of both: embracing new technologies where they enhance educational experiences, while maintaining the standards and fairness of tried and tested assessment methods. 
• Exams are often the best way to gauge what students have learnt. The success of the return to in-person exams in 2022 demonstrates the importance of well-established, human-centred and continuously evolving assessment tools - especially as new technologies change how we teach and learn. 

We are supporting International schools with guidance and we've shared Joint Council for Qualifications guidance for schools and colleges in the UK. 

Cambridge expertise

Our combination of deep education, research and technological experience and expertise means we are well-placed to help schools, colleges, teachers and other partners benefit from the opportunities of this rapidly emerging technology.

Cambridge has been using AI for years in appropriate circumstances, including to accelerate results and provide quality control for some exams, like Linguaskill’s English language assessments, which use AI in scoring and automarking – augmenting human expertise. We are continuing to innovate and experiment in the use of AI to improve education.  

Beyond assessment, we have already published an ethics policy for AI in research publishing.