Skip to content

AI-use guidance for term papers, theses and dissertations: rule files, guides and log templates (German/English)

Sep 2026 · Zenodo (CERN European Organization for Nuclear Research)

Abstract

English: (German below) A bilingual (German/English) set of materials for using generative AI in academic term papers, theses and dissertations. A rule file, pasted once into a chat project, turns a general chatbot into a tutor that explains rather than supplies, questions rather than agrees, and declines requests that would substitute for the student's own writing, analysis or source work. Requests are sorted into green, amber and red; material from assignments that permit no AI help is refused outright. Brainstorming is available as a named, student-triggered mode; and the assistant keeps a running AI-use log that the student completes and keeps. The guide explains the reasoning behind every rule, how to set the rules up and test them, how to use AI in the research and writing process, and what the student's own university regulations take precedence over. This guide was written originally for students of German and English linguistics at Ruhr-Universität Bochum and Humboldt-Universität zu Berlin. The current version is adapted to the dominant models available as of late September 2026; I intend to update this guide and repository at least biannually, as the field develops and as needs change. Licensed under CC BY 4.0 for reuse and adaptation with attribution. The live versions can be found here: German: https://amor.cms.hu-berlin.de/~loslucbe/ki-leitfaden.htmlEnglish: https://amor.cms.hu-berlin.de/~loslucbe/ai-guide.html Ein zweisprachiges (Deutsch/Englisch) Materialpaket für den Einsatz generativer KI in akademischen Hausarbeiten, Abschlussarbeiten und Dissertationen. Eine Regeldatei, die einmalig in ein Chat-Projekt eingefügt wird, verwandelt einen allgemeinen Chatbot in einen Tutor, der erklärt statt nur Informationen zu liefern, Fragen stellt statt einfach zuzustimmen und Anfragen ablehnt, die das eigene Schreiben, die Analyse oder die Quellenarbeit des Studierenden ersetzen würden. Anfragen werden in grün, gelb und rot sortiert; Material aus Aufgaben, bei denen keine KI-Hilfe erlaubt ist, wird sofort abgelehnt. Brainstorming steht als benannter, vom Studierenden ausgelöster Modus zur Verfügung; und der Assistent führt ein fortlaufendes KI-Nutzungsprotokoll, das der Studierende vervollständigt und aufbewahrt. Der Leitfaden erläutert die Gründe für jede Regel, wie die Regeln eingerichtet und getestet werden, wie KI im Recherche- und Schreibprozess eingesetzt wird und welche Bestimmungen der eigenen Hochschule Vorrang haben. Dieser Leitfaden wurde ursprünglich für Studierende der germanistischen und anglistischen Linguistik an der Ruhr-Universität Bochum und der Humboldt-Universität zu Berlin verfasst. Die aktuelle Version ist an die gängigen Modelle angepasst, die seit Ende September 2026 verfügbar sind; ich beabsichtige, diesen Leitfaden und das Repository mindestens zweimal jährlich zu aktualisieren, entsprechend der Entwicklung des Fachgebiets und den sich ändernden Anforderungen. Lizenziert unter CC BY 4.0 zur Weiterverwendung und Bearbeitung mit Namensnennung. Use of AI in producing this resource: the guides and rule files were drafted and revised with the help of ChatGPT (OpenAI) and Claude (Anthropic) based on rigorous testing of the rule set with automated runs and manual testing across OpenAI and Anthropic models. These tests checked behaviour in a limited set of cases, proving that the established guardrails are functional. However, the testing does not mean that the designed guardrails cannot be jailbroken.

View source

Similar papers

#artificial intelligence Conference Open access Apr 2020

ECCOLA - a Method for Implementing Ethically Aligned AI Systems

The method, ECCOLA, is presented, which aims at making the high-level AI ethics principles more practical, making it possible for developers to more easily implement them in practice.

Ville Vakkuri, Kai-Kristian Kemell, P. Abrahamsson · 64 citations · ⚡6
#computer vision Review Apr 2024

AI-powered Code Review with LLMs: Early Results

The goal is to not only refine the accuracy of the LLM-based tool but also to underscore its potential in streamlining the software development lifecycle through proactive code improvement and education.

Z. Rasheed, Malik Abdul Sami, Muhammad Waseem et al. · 62 citations · ⚡3
#computer vision Open access Mar 2024

LLM-based agents for automating the enhancement of user story quality: An early report

The use of large language models to automatically improve the user story quality in Austrian Post Group IT agile teams is explored, with a reference model for an Autonomous LLM-based Agent System developed and implemented at the company.

Zheying Zhang, M. Rayhan, Tomas Herda et al. · 48 citations · ⚡4
#computer vision Review Mar 2024

System for systematic literature review using multiple AI agents: Concept and an empirical evaluation

This paper introduces a novel multi-AI-agent system designed to fully automate SLRs, and demonstrates how it substantially reduces the time and effort traditionally required for SLRs while maintaining comprehensiveness and precision.

Abdul Malik Sami, Z. Rasheed, Kai-Kristian Kemell et al. · 44 citations · ⚡2
#computer vision Feb 2024

Can Large Language Models Serve as Data Analysts? A Multi-Agent Assisted Approach for Qualitative Data Analysis

The proposed LLM-based multi-agent system automates qualitative data analysis process, creating opportunities for researchers and practitioners, and future improvements focus on enhancing multilingual performance and integrating continuous expert feedback.

Z. Rasheed, Muhammad Waseem, Aakash Ahmad et al. · 41 citations

Related blog posts

Microsoft Research Blog Oct 7, 2026

Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses

Training AI agents with reinforcement learning can be challenging because their tools, context, and decision-making are managed by complex frameworks. Agent Lightning connects existing agents to RL training, making it easier to improve them without rebuilding them. The post Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.