AI is a collective term for technologies that can mimic human behavior and thinking.

Artificial intelligence is information technology that adjusts its own activity and therefore seemingly appears intelligent. (Translated from: Store norske leksikon)

AI is not a single technology, but a set of technologies that use large amounts of data, complex algorithms, and powerful computational capacity to perform tasks with little or no human guidance.

Generative AI and its possibilities

Generative AI tools, such as Microsoft Copilot, can create text and images, perform analyses, and write code. While this is impressive, it is important to remember that AI does not "understand" content in the same way humans do. It recognizes patterns and generates responses based on probabilities, not facts.

Critical thinking: AI is not infallible

AI tools are useful, but they are not infallible. AI cannot be held accountable for the answers/products they generate and should not make decisions. It is humans who must take responsibility for the quality and be critical of the results generated.

Language models: What they can and cannot do

When you ask a question or give a task to a chatbot, it uses the patterns it has discovered in its training data to predict which words statistically fit well together as a response to your question. Language models can generate text that seems natural, but they do not have fact-based understanding. They are good at simulating conversations but can provide incorrect answers without being aware of it.

AI tools connected to academic databases

Some AI tools are connected to academic databases and provide answers that include citations. However, be aware that not everything in academic databases is of equally high quality, and AI cannot assess the reliability of sources. You still need to evaluate the quality of the information yourself.

Microsoft Copilot is available for students and employees at NTNU.

Microsoft Copilot is an AI chat similar to other well-known tools such as ChatGPT. It is a tool integrated into your browser, and you can get help with tasks such as text improvement, image generation, performing analyses, and writing code."

What happens when you use an AI chat?

AI chat predicts responses based on patterns in training data. They do not search for facts but generate text that statistically seems correct. This means that:

• They can provide incorrect answers in a convincing manner.
• They do not have source criticism or fact-checking.
• They can "hallucinate" and invent information.
  • When AI generates text, it can sometimes create information that seems credible but is actually incorrect or unverified. This is called "hallucination" – erroneous but seemingly credible information.

Speech to text  
This service uses artificial intelligence to automatically transcribe audio and video files, both with open (green) and internal (yellow) data. It can handle various file types, including recordings from Zoom and Panopto, and streamline video captioning.

Nettskjema-diktafon
The Nettskjema Dictaphone mobile app can be used to make audio recordings of interviews, with automatic transcription using the Autotekst tool. Nettskjema is a service NTNU purchases from the University of Oslo.

Microsoft Copilot - can be used to generate images.

When using AI for image processing, remember the following:

• Do not generate images or videos of others without consent. Images that show people are personal data and are covered by the General Data Protection Regulation (GDPR) (in Norwegian).
• AI-generated images and videos may also be subject to copyright. You are therefore obliged to familiarise yourself with the documentation from the tool on how the service handles copyright.
• Label AI-generated images/videos clearly to indicate that they were created using AI and specify which tools were used.

Example:

Scopus AI is a technology in the Scopus database that uses artificial intelligence to help you find relevant articles, identify trends, and gain insights into your field faster and more accurately. Instead of a list of hits, the result is a summary with references to academic sources indexed in the Scopus database

If you are not on campus, you must install VPN to get access.

NTNU has a license for Scopus AI until 2025 (as a trial period).

We encourage you to use NTNU services whenever possible.

However, there are many commercial AI tools available online, and it is easy to create AI tools yourself. These are mainly tools designed for private and personal use. It is permissible to use commercial AI tools at NTNU, but you should only share open information, and personal data should not be included in instructions to an AI tool. Use of commercial AI tools for other purposes must be in accordance with other guidelines in the information security management system.

When you use an AI tool that NTNU has not approved, you are sharing information with external parties. This can be both unfortunate and illegal, depending on the information you provide. These tools should only be used for general questions. In such AI tools, you should:

• Never share personal information - neither about yourself nor others.
• Never share confidential information or information that is exempt from public disclosure.
• Never share login information or passwords.
• Do not paste content from documents that are exempt from public disclosure.

Important to be aware of:

• Instructions you give are stored somewhere you do not have control over. It will be nearly impossible to access, retrieve, or delete the instructions.
• If you make personal data available about others than yourself, you are violating their privacy rights.
• Instructions you give may become part of the language model's training data. This may be made available to other users of the tool or used for other purposes.
• Instructions you give may be made available as publicly accessible data or sold. This includes files or other content you include in the instructions.

Teaching and Exams

If an teacher requires students to use AI tools in teaching or in connection with exams, tools approved by NTNU must be used. This means that the tool must either be developed by NTNU or that NTNU has entered into a data processing agreement with the provider. NTNU cannot require students to use AI tools other than those approved by NTNU.

NTNU has its own website for

Artificial intelligence in student papers

It is permitted to use AI as an aid in preparing student assignments, but the use of AI may vary between subjects and study levels. Ask your subject teacher or supervisor if you are unsure about AI use. It is your responsibility to act in an academically honest manner and to familiarize yourself with applicable ethical principles, laws and regulations.

NTNU's guide to using generative AI in research

• Researchers are responsible for their own research results.
• Be open about the use of generative AI.
• Pay particular attention to issues related to privacy, confidentiality and intellectual property rights when using AI tools.
• Respect applicable national, EU and international legislation, as in all research.
• Keep up to date on how to use generative AI tools to exploit their benefits, and be aware of disadvantages and limitations.
• Avoid using generative AI tools in sensitive activities that may influence other researchers or organizations, such as peer review or evaluation of applications.

The guide is based on, and is updated in accordance with, Living guidelines on the responsible use of generative AI in research.

There are many areas where you as a student can benefit from artificial intelligence. Here are some examples:

• Assignment planning: Get help structuring academic assignments.
• Proofreading and grammar help: Use AI tools to support your work in writing academic texts.
• Brainstorming: Get ideas for assignments or projects.
• Summaries:AI can support reading, including by creating summaries of long or complicated texts.
• Improve academic style: AI can support you in getting a consistent and correct style in the text, by giving you feedback that can support you in your writing work.
• Remove bias in texts: Use AI to identify and reduce bias in writing.
• Task list and planning:Use AI to create effective study plans.
• Quiz generation: Create questions based on the syllabus to test yourself.
• Translation: Get texts translated - here we also have other services, currently Word translates well, but AI can also give you multiple levels of explanations of technical terms or linguistic expressions.

It is important to be critical of the answers AI tools provide and always double-check the information before using it. See more about this under principles and guidelines.  

AI tools can be useful in your studies, but they can also provide incorrect or misleading answers.

It is important to:

• use AI as support, not as a substitute for your own thinking.
• critically assess the quality of the answers you receive.
• always disclose the use of AI in academic work and clearly label material generated by language models.
• do not use AI to process personal data or sensitive matters.
• follow professional and ethical guidelines for Using and citing sources.

• Writing support: Use AI tools like Microsoft Copilot to improve the language of your academic texts, such as grammar and style, but be sure to acknowledge that you have used AI.
• Idea generation: Use AI to generate ideas or suggestions for the structure of a paper, but develop and write the content yourself.
• Data analysis: Some tools with AI technology can be suitable for data analysis. It is important that the analyses follow good academic practice and have a high level of integrity.

• Plagiarism: Do not submit texts generated by AI, where you do not have ownership of the formulation and have not assessed it yourself. Be critical of the sources you use and refer to sources of information and knowledge you use in the assignment. Read more about plagiarism.
• False information: Do not use AI to generate or spread incorrect information or data that cannot be verified.
• Sharing of personal information and sensitive data: AI tools should not be used in a way that involves sharing sensitive personal information or data that is subject to confidentiality or legal restrictions. Examples of this are:
  • Personal data (e.g. names, addresses, student numbers)
  • Sensitive research data or data that can be linked to individuals (e.g. health information or financial data)
  • Data subject to confidentiality or other legal restrictions.
  • Images where individual individuals can be identified

Read more about collection of personal data for research projects

NTNU uses the Data storage guide to define which data can be used in our AI services. The data is divided into four categories to ensure that the information is handled in a safe and correct manner:

🟢 Open (Green data)
Open information that is available to everyone without special access rights.

🟡 Internal (Yellow data)
Information that is limited to internal use and is only available to employees who need it to perform their tasks.

🔴 Confidential (Red data)
Information that, if made known to unauthorized parties, could harm public interests, NTNU, individuals or partners.

⚫ Strictly confidential (Black data)
Information that, if it becomes known to unauthorized persons, may cause significant damage to public interests, NTNU, individuals or partners.

Use of data in tools and services

Rules for which data can be used in different tools and services vary, depending on assessments of information security and privacy. Before using a tool, it is important that you familiarize yourself with the guidelines for the tool in question, as the use of data may be regulated.

Assessment of Tools

It is necessary to carefully assess the tools before they are used. For example, NTNU has conducted thorough assessments of Microsoft Copilot. After assessing information security and privacy, NTNU has concluded that this tool can be used for data classified as Open (Green data) and Internal (Yellow data).

Important: Always remember to check the tool's guidelines before use, to ensure that you are using the data in a safe and correct manner.

• The Personal Data Act (GDPR) - protects personal data and gives individuals control over their data. It requires that data is processed securely and lawfully.
• AI Act -  EU law that regulates the use of artificial intelligence, to ensure that AI is used in a safe and ethical manner.
• Public Administration Act - Regulates how public institutions (such as NTNU) make decisions, and ensures fair and open handling of cases.
• Freedom of Information Act - Gir allmennheten rett til innsyn i offentlige dokumenter, men beskytter sensitive opplysninger.

Policy for the use of ICT tools with generative artificial intelligence at NTNU

• Nasjonal strategi for KI (Regjeringen) (in Norwegian)
• Retningslinjer (Teknologirådet) (in Norwegian)
• Samling ressurser KI (Digdir) (in Norwegian)

• Elements of AI - digital course
• Explore the study program in artificial intelligence
• AI Seminar series - Physically at NTNU (open to all)
• Maskiner som tenker - Book. Author: Inga Strumke. Associate Professor, NTNU

• Artificial Intelligence at NTNU 
• Using artificial intelligence as an employee