十校政策记录与证据修订

本次修订统一现行页面、大学 JSON、工具资料与 MCP；清理引文污染、错误引用及审核状态说明。旧基线、旧批准与暂缓条款单独保留；这是网站记录更正，不代表大学刚颁布新政策。

历史修订：以下记录数与主张集合以本版发布时为准，现行记录另见最新修订。

2026-10-14T02:23:21.633198Z · v6-audit-fix-20261014

• Imperial College London · 17 条现行审核记录 · 查看政策页
  • 新增记录: 0; 更正记录: 0; 历史／暂缓: 0

查看上一版 V6 发布记录

记录修订、适用范围与原文

原文与人员、课程、学部范围保留。网站修订时间不是大学政策生效时间。

f42affcf950f80c77f48d2cf1ee59abc09f156edfbb7147f8c03864bf2fb2211

{
  "addedClaimIds": [],
  "updatedClaimIds": [],
  "heldClaimIds": []
}

• cl-imperial-college-london-chatgpt-14 — Imperial provides access to selected OpenAI GPT models through the institutional dAIsy platform. This does not establish access to the standalone ChatGPT service or permission to use GPT-generated content in assessed work.
• cl-imperial-college-london-claude-15 — Imperial provides access to selected Anthropic Claude models through the institutional dAIsy platform. This does not establish permission to use Claude-generated content in assessed work.
• cl-imperial-college-london-deepseek-16 — Imperial provides access to selected DeepSeek models through the institutional dAIsy platform. This does not establish permission to use DeepSeek-generated content in assessed work.
• cl-imperial-college-london-institutional-ai-service-17 — Imperial provides dAIsy as its institutional multi-model generative-AI platform. Basic access is available to eligible staff and students to create Personal Agents; Developer access with shared agents is restricted to approved staff, and direct LLM API access is not currently provided. Platform access does not constitute coursework permission.
• clm-imperial-college-london-ai-misconduct — Unless explicitly authorised, using artificial intelligence to create assessed work may be treated as an offence such as contract cheating under Imperial's Plagiarism, Academic Integrity & Exam Offences regulations.
• clm-imperial-college-london-dept-assessment-policy — Individual Imperial departments may allow or prohibit generative AI for specific assessments; students must follow the assessment brief and local instructions. The dAIsy Use Policy complements existing university policies and does not replace them; university policy prevails where there is a conflict.
• clm-imperial-college-london-daisy-privacy — Imperial's dAIsy AI platform uses University SSO authentication with auditing. Prompts and metadata are logged for operational monitoring, and AI model providers are configured not to train on user data. Users' prompts and responses are not used to train external AI models. dAIsy is approved for use with unrestricted data within Imperial's secure infrastructure.
• clm-imperial-college-london-daisy-sanctions — Breaches of Imperial's dAIsy Use Policy may lead to action under Academic Misconduct procedures for students and HR/disciplinary processes for staff, as well as under Information Security and Data Protection policies. Sanctions may include removal of access, grade penalties, or formal disciplinary measures.
• clm-imperial-college-london-ai-acknowledgment — Imperial Library guidance advises students to acknowledge generative AI use in assessed work and outlines recommended elements. When dAIsy is permitted for assessed work, its Use Policy requires students to acknowledge AI use and maintain transparency. Specific requirements and formats are determined by departments, courses, or assessment briefs.
• clm-imperial-college-london-research-ethics — Research at Imperial that involves people, personal data, or sensitive topics may require ethics approval, a Data Protection Impact Assessment (DPIA), and data-governance controls before using any AI tool. Researchers must verify whether their use of AI in research requires special approval, particularly when uploading private or confidential research data.
• clm-imperial-college-london-copilot-data-protection — Imperial provides Microsoft Copilot access to staff and students using institutional accounts. Microsoft Copilot Chat offers enterprise data protection, and prompts and responses are not used to train foundation models. Microsoft documents that Copilot Chat prompts and responses are logged and stored in Exchange within the Microsoft 365 tenant for auditing and eDiscovery. Access to organisational information and specific retention controls depend on the Microsoft product, licence, and tenant configuration. Users should follow Imperial's current data-protection requirements. When web grounding is enabled, generated search queries can be sent to Bing, which operates separately from Microsoft 365 under different data-handling terms. This record does not establish which features are enabled in Imperial's tenant.
• clm-imperial-college-london-user-accountability — Users of Imperial's dAIsy AI platform must always apply critical judgment to AI outputs. Generative AI can produce inaccurate or biased outputs ('hallucinate'), omit context, or reflect training-data biases. Users remain accountable for the accuracy, legality, and appropriateness of any content they submit or share through the platform.
• clm-imperial-college-london-daisy-copyright-ip — Users of Imperial's dAIsy platform must not upload third-party content they are not permitted to share. Reuse of AI outputs must comply with licensing and academic citation norms. When communicating externally, dAIsy outputs must not be presented as Imperial's position without approval.
• clm-imperial-college-london-genai-principles — Imperial's generative AI guidance presents three areas of principles for teaching, learning and assessment: critical use of AI, a consistent ethical approach, and a proactive research community. They are underpinned by five core Values: Respect, Collaboration, Excellence, Integrity and Innovation. The principles are a non-exhaustive foundational framework; they do not establish blanket assessment permission.
• clm-imperial-college-london-ai-detection-2026 — For the 2026–27 academic year, Imperial states that it does not use Turnitin's AI-writing detection functionality as part of its assessment processes. Academic-integrity concerns are evaluated by staff using academic judgement and evidence. This statement does not exclude authenticity interviews or other permitted investigative methods.
• clm-imperial-college-london-turnitin-ai-data-terms-2026 — Imperial states that Turnitin's proposed July 2026 changes concerning AI-related processing of student submissions have been deferred until at least the 2027–28 academic year. The university reports no corresponding change to its Turnitin usage arrangements during 2026–27.
• clm-imperial-college-london-academic-misconduct-appeals — Imperial's Academic Misconduct procedures govern the investigation of suspected misconduct, potential sanctions, and applicable appeal rights. Students challenging an Academic Misconduct Panel decision must submit an appeal within ten College working days of notification; the general academic-results appeal procedure (with a 15 working-day window) does not apply to misconduct decisions. Potential misconduct in postgraduate research programmes is dealt with under Ordinance D17; the ten-day statement does not establish a deadline for that separate procedure.