AI In Education

Generative AI, as defined by UNESCO, is an artificial intelligence technology that generates content like text, images, code, or audio in response to natural language inputs, using huge amounts of data it was trained on.

This technology offers accessible tools with broad applications for both students and teachers, which could impact how students study and how teachers teach. For example, with students being able to use AI to write essays, teachers may worry that this assessment method is no longer reliable. This example and other examples of AI implications for education are further discussed in our resource hub.

Students and employees within the University of Twente are able to use Microsoft Copilot by logging in with their UT credentials. However: while this version of Copilot is included within UT’s Microsoft license, it does not (yet) adhere to the privacy agreements in data processing for the University of Twente. It’s not allowed to use any confidential data like personal, research or student information in this. SURF actively advises against the use of Copilot.

Besides this, it’s possible to run a LLM on your own machine. Open source software like Ollama and LM Studio allow you to run any open source model locally, and models with a lower amount of parameters do run on regular UT devices. 

To learn more, visit the resource hub section Fundamentals: what tech is out there, what to keep in mind.

Some general guidelines for safe and responsible use that may apply to all users of AI tools:

• Don’t use personal or confidential information in unlicensed online tools. Your data may be stored, sold and used to train future models.
• Verify the accuracy of AI-generated content to prevent the spread of information that is incorrect, unfair or biased.
• Be transparent and disclose when content is created by AI to maintain trust.
• Ensure AI output does not infringe on copyrights by checking for potential replication of training data.
• Consider the environmental impact, as running large AI models can be resource-intensive.
• View AI output as a starting point, not a final answer.

To learn more, visit the resource hub section Fundamentals: what tech is out there, what to keep in mind.

There are various ways for teachers to use AI to assist them in their tasks. For example: using AI to assist you in designing a course or creating course materials. To learn more, visit the resource hub section dedicated to the topic of AI as a teaching assistant.

As a programme developer, you may consider various factors to make this decision, including:

• AI use in the graduates’ future work field
• The fact that the skilful use of AI to search and filter information is considered part of the necessary digital competencies for all EU citizens (cf. DigiComp 2.2)
• The importance of ensuring graduates’ awareness of ethical concerns and fostering their development of a human-centric mindset in interacting with AI (cf. UNESCO AI competency framework for students)

From a practical perspective, integrating AI literacy does not necessarily equal a complete overhaul of the curriculum. There are various opportunities to integrate AI literacy into an existing curriculum without detracting from the existing content and even achieving benefits such as student engagement, personalised learning, and the achievement of learning outcomes.

To learn more, visit the resource hub section dedicated to the topic AI literacy in the curriculum.

Several strategies are possible:

(1) You may integrate AI-related information (e.g., related to the potential, limitations, and ethical concerns of AI) in your classes and accompany them with learning activities for deeper engagement.

(2) You may present students with learning activities designed to stimulate the development of higher-order skills and/or domain knowledge while simultaneously fostering AI literacy.

(3) You may provide students with activities using AI as a learning assistant, for them to learn about AI while reaping the benefits of personalised learning and student engagement.

Do you want to learn more or put this in practice? See the resources and examples on the resource hub’s section on ‘AI literacy in courses’.

As a teacher, you must specify what type of use of AI in assessment is authorised or not in your course AI policy. However, since AI detectors are not fully accurate, students may be tempted to make use of AI in assessment products even if unauthorised. There are two things you can do:

(1) Be clear (in your course AI policy and in conversation with your students) about the fact that unauthorised AI use is regarded as fraud and set strict measures in case of suspicions acting as a deterrent.

(2) Consider your assessment method and whether your ILOs require you to assess your students in an AI-free environment, change assessment method, redesign the assessment tasks, or even if incorporating AI into assessment may be beneficial.

You can find more information, examples, and guidance on the resource hub section on AI in assessment.

Absent a faculty- or programme-level AI policy, every teacher must define a policy specific to their course. The policy must specify:

• Whether AI is allowed in learning and, particularly, assessment tasks
• Which categories of AI tools are authorised and in what ways
• How students should declare the use of AI in assessment products, and
• What consequences apply in case of unauthorised use.

Visit the section of the resource hub on ‘AI and assessment’ to get a template for your course-specific AI policy and guidance on how to deal with suspected AI use.

You may NOT use AI to fully automate the grading of student work. However, you may use AI tools to increase the efficiency of YOUR grading if such use complies with the requirements for safe and responsible use (EU AI Act, GDPR).

For more information and examples, visit the section of the resource hub on AI and Assessment.