King's College London

King's College London does not ban the use of generative AI by students. Whether and how AI may be used in a given assessment is set by the assessment guidance: students should be given clear guidance on the level of AI that is appropriate and should ask if uncertain.

King's Academic Misconduct Policy states that using generative AI tools to produce output that is then copied as part of a submission is not appropriate, that any work submitted must genuinely demonstrate the student's own work, skills and knowledge and follow the assessment guidelines, and that generative AI writing programmes are third-party technologies to be used only in accordance with College guidance. The King's guidance adds that submitting AI-generated text as one's own without written departmental permission is considered misconduct (third-party involvement or text manipulation) where it gives undue advantage or interferes with assessment of the student's own understanding. A written permission or acknowledgement does not by itself make substantive AI content creation acceptable.

King's College London suggests four broad levels of generative AI use in assessment that programme and module leaders may adjust to each assessment: Minimal (routine tools such as auto transcription, spell and grammar checkers), Limited/selective (clearly delineated tasks as allowed), Open (no specific restrictions but key stages and tools must be tracked) and Embedded (AI use is a focal part of the assessment). The levels are a framework for the actual assessment brief, not an automatic university-wide permission, and verbatim use of generated content without acknowledgement is never appropriate.

King's College London does not require students to cite generative AI as an authoritative source in the reference list, but it requires students to acknowledge any use of generative AI tools in coursework by including a declaration statement with their references. Unless the assessment brief gives other wording, students append either a declaration that no part of the submission was generated by AI software (English grammar and spell checking is consistent with this statement) or a declaration that parts of the submission have AI contributions within the acceptable use specified in the brief, naming the tools and how they were used.

King's College London states that its Enterprise licence means Copilot use comes with commercial data protection, making it a more secure alternative to other generative AI tools. Commercial data protection does not mean any data may be entered: King's student guidance says not to input personal data, sensitive information or content created by others into any AI platform.

King's College London's student guidance states that Microsoft Copilot is available to all King's students with their KCL Microsoft login credentials. The guidance does not describe Copilot as the only, primary or required AI tool, and availability does not grant permission to use it in a specific assessment.

King's College London states that all staff and students have access to text and image generation through Microsoft Copilot when logged into their KCL Microsoft account, and that its Enterprise licence provides commercial data protection. This is tool access only: it does not authorise AI use in a particular assessment, and the student guidance separately tells students not to input personal data, sensitive information or content created by others into any AI platform.

King's College London permits doctoral students to use generative AI tools in an assistive role in thesis writing, to clarify their own writing and avoid misrepresenting their meaning; this can include rewriting, rephrasing or paraphrasing parts of the thesis. Use beyond the circumstances permitted by the doctoral guidance risks breaching the Academic Misconduct Policy, and AI used in the research process itself (for example to analyse data) is largely outside this guidance and falls under research design, ethical review and research-integrity norms.

King's College London states it took the decision not to enable the AI-writing detection percentage in Turnitin because of concerns about reliability and false positives. This is not the same as switching off similarity checking: the Academic Misconduct Policy still allows staff to use TurnitinUK or other means to help detect academic misconduct. The current back-end Turnitin configuration was not independently tested.

King's College London states that it contributed to and subscribes to the Russell Group's five principles on the use of generative AI tools in education, which frame its governance and AI-literacy work. These are principles, not a course-level permission.