AI — FAQ

Article sections
Below you can find answers to some of the most frequently asked questions at HAMK, particularly concerning generative AI.
1. What are the most important things to consider when using AI?
- AI is a support tool: AI can be used, for example, for brainstorming, planning, problem-solving and routine tasks. Using AI requires your own judgement: check essential information and sources, and assess the reliability and potential biases of AI-generated content. Responsibility for the use of AI and for decisions made with its assistance remains with people and the organisation according to their respective roles and responsibilities. See also Ethical principles for the use of AI in brief – Digital Pedagogy Guidelines. HAMK’s ethical principles emphasise fairness, transparency, human oversight, privacy and responsible use of AI.
- Guidance for students: Plan the use of AI according to the learning objectives so that it supports the development of students’ own thinking, competence and working-life skills. In the assignment instructions, explain how AI may be used, how its use should be reported and how it will be taken into account in assessment. Use the AI traffic light model when appropriate. You should also be transparent about your own use of AI to support guidance, feedback and assessment. See Examples of using Arene’s AI traffic light model – Digital Pedagogy Guidelines. HAMK’s guidance specifies that the traffic light category should be selected according to the learning objective and that students should be told how AI use affects assessment.
- Data protection: Follow HAMK’s information classification and data protection guidelines, and primarily use AI services provided and approved by HAMK. Personal data, students’ learning outputs, and confidential business and research materials may only be processed when there is an appropriate legal basis for doing so and the service has been approved for processing the relevant data. Also make sure that you have the necessary copyright and licensing rights to use the material. The fact that material is not used to train an AI model is not, on its own, sufficient to determine whether processing the material in a service is acceptable. See Information classification – Digital Pedagogy Guidelines. HAMK’s information classification defines different categories ranging from Secret (1R) to Open (5W), with different handling requirements.
2. How do I get started with AI and develop my AI skills?
- The best way to get started is to experiment with AI and learn by using it. It is also worth asking colleagues for tips and sharing ideas with them.
- Next, complete the AI learning module in the Staff Digital Skills Package and the HAMK AI Skills badges. HAMK currently provides three AI skills badges: Basic AI User, Pedagogical AI Expert and AI Citizen.
- We also recommend following discussions and tips in the Teams- channel, and participating in HAMK’s AI Coffee sessions and training.
- EduHouse offers a free training programme on using ChatGPT: ChatGPT – tee työtä tekoälyn kanssa.
3. What generative AI applications are commonly used at HAMK?
- Microsoft Copilot: Microsoft’s AI assistant, which can be used for purposes such as searching for information, creating content and summarising web pages. In the Edge browser, Copilot can help you work with content that is open in the browser. For work and study, Microsoft 365 Copilot Chat and the Microsoft 365 Copilot features available with particular licences provide different ways of working with organisational data.
- Gemini: Google’s generative AI assistant, which can help with writing, brainstorming, learning and summarising materials, for example. It uses multimodal AI, which can process not only text but also content such as images and audio. The service was previously known as Bard.
- Gemini Notebook (formerly NotebookLM): Google’s source-based AI assistant designed to support research and learning. It allows you to ask questions about source materials and create summaries, audio and video overviews, concept maps, infographics and presentations, for example. Source citations included in the responses help you verify information against the original materials.
4. Can I sign in to AI applications using my HAMK account?
- Microsoft Copilot Chat: You can sign in using your HAMK Microsoft account. When you use the service with your organisational account, enterprise data protection applies, and your prompts and responses are not used to train the underlying AI models. This also applies to Copilot Chat in the Edge browser when used with your organisational account.
- Gemini: You can sign in using your HAMK Google account. When using the service with HAMK’s Google Workspace for Education account, your conversations and uploaded files are not used to train generative AI models. Make sure you are using your HAMK account rather than a personal Google account.
- Gemini Notebook (formerly NotebookLM): You can sign in using your HAMK Google account. When the service is used with HAMK’s Google Workspace for Education account, your materials are not used to train AI models. Again, make sure you are using your HAMK account.
- Signing in with your HAMK account does not, on its own, mean that any type of information may be processed in the service. Always follow HAMK’s information classification requirements and service-specific instructions. HAMK’s information classification guidance defines how information of different classifications must be handled.
5. Which AI applications can I use to create images?
- Adobe Firefly: Adobe’s AI service for generating and editing images. Adobe’s own Firefly models have been trained using licensed and copyright-free content. According to Adobe, content uploaded to or generated in the service is not used to train its AI models. Note that the service also includes models from other providers.
- Microsoft Copilot: Generates images from text descriptions, for example for teaching materials and presentations. Images can be created in Copilot Chat or using the Create feature. Availability depends on your licence and your organisation’s settings.
- Google Gemini: Can generate new images and edit existing ones, for example by replacing backgrounds or adding and removing elements. Some features may be restricted for educational accounts.
6. I would like to get Microsoft 365 Copilot. What should I do?
Copilot Chat is available with an eligible Microsoft 365 account without a separate Copilot licence. The broader Microsoft 365 Copilot features require a paid licence, which may appear in the service as M365 Copilot (Premium).
- Complete the Basic AI User skills badge through the HAMK AI Skills page. Completion of this badge is required if you want access to a paid AI application such as Microsoft 365 Copilot.
- Complete the application form in the HAMK Helpdesk. You can find the form under IT services and devices → Software, licences and settings.
7. I would like to use a paid AI application. What should I do?
Submit a request to the HAMK Helpdesk so that your needs can be assessed and the procurement process can be initiated.
To obtain a paid AI application, you must first complete the Basic AI User skills badge through the HAMK AI Skills page. HAMK’s internal skills page confirms that completion of the badge is required for access to paid AI applications such as Microsoft 365 Copilot.
See also Guidance for users testing an AI application – Digital Pedagogy Guidelines. HAMK recommends primarily using AI tools provided and approved by the organisation.
8. How can I use AI while respecting copyright?
Only use materials in an AI service if you have the necessary rights to use them. Do not upload someone else’s copyrighted material unless you have permission or a licence that allows you to use it in this way.
Check the terms of use for the material and review AI-generated content before sharing it. HAMK’s Digital Pedagogy Guidelines recommend using appropriately licensed or open materials and complying with the applicable licence conditions.
See also How can I use AI while respecting copyright? – Digital Pedagogy Guidelines.
9. How much energy and water does AI use?
There is no single universally applicable figure for the energy consumption or water use of AI. Consumption depends on factors such as the model used, the task, the hardware and data centre cooling. The scope of the calculation also matters: generating an individual response is different from assessing the overall footprint that includes model training and hardware manufacturing.
The consumption of an individual text prompt can be small. For example, in a measurement published by Google in 2025, the median energy consumption of a Gemini Apps text prompt was 0.24 watt-hours of electricity and 0.26 millilitres of water, approximately five drops of water. These figures relate to that particular service and measurement period and should not be generalised to all AI services.
Generating images, and especially videos, can consume considerably more energy than generating text. Consumption varies depending on the model and settings. Video length, resolution and generation settings affect energy requirements, and every new generation increases overall consumption.
Water is used both directly for data centre cooling and indirectly in electricity generation. Some cooling water evaporates, so recycling water does not eliminate water consumption. When evaluating environmental impacts, the availability of local water resources should also be taken into account.
Improvements in energy efficiency can reduce consumption per individual task, but increasing use may still increase total consumption.
Recommendation for users: use AI purposefully, and generate images, videos and new versions of them when there is a genuine need.
