Adelina Silva teaches Technology Education in lower secondary (9th grade), and has progressively been integrating digital literacy and Artificial Intelligence into the lessons. Her focus is always on developing critical thinking and ethical responsibility in adopting generative AI. In this unit she connects science and citizenship, preparing her 9th-grade students for a future where AI is present, teaching them to use these tools, to question them, to verify the information they generate, and to act responsibly.

Part One – Introduction and engagement

Adelina opened the lesson with a 5-minute ice-breaker, asking: “How many energy transformations happened between the moment you woke up and arriving at school?”

Students share their answers. The teacher uses them to introduce the law of conservation of energy and the types of energy (kinetic, potential, thermal, electrical, among others).

Activity 1

Each student fills in a concept map on energy with the concepts they already know, in Section A of Worksheet Lesson 1.

Part Two – Exploration and application

The teacher explains the next activity, which links what they know about the law of conservation of energy with the critical evaluation of AI.

Activity 2

In pairs, students choose a generative AI tool to work with (ChatGPT or Copilot). In the chosen tool:

• they type the questions indicated in Section B of the Worksheet Lesson 1: “What is energy? What types of energy exist and how do they transform?”, without prior guidance from the teacher.
• Filling in Section B of Worksheet Lesson 1:

“AI’s summarized answer”; “I agree or disagree and why”; “What I need to verify”.

Activity 3: Plenary discussion (5 min)

The teacher creates a plenary discussion (5 min) where all students can share:

Did the AI answer well? How do we know? What criteria do we use to assess the quality of an answer?

The aim is to start thinking about ethical and academic-integrity questions, even if indirectly.

Activity 4: Key concepts about AI

Next, the teacher briefly introduces some concepts that help students better understand how AI technology works:

• The concept of AI hallucination and why it happens.
• The types of AI (narrow AI vs general AI), with everyday examples such as Netflix recommendations, email spam filters, automatic text translation.
• She explains what generative AI is, what large language models (LLMs) are, and gives examples of tools such as ChatGPT.

Activity 5: Class Ethics Contract (20 min)

The teacher invites students to bring together the experience, reflection and knowledge they now have about generative AI, and to create the Class Ethics Contract for the use of generative AI.

In heterogeneous groups of 4 students, the class creates the Ethics Contract for the use of generative AI:

• Each group discusses and proposes 2–3 rules, filling in Section C Worksheet Lesson 1.
• The teacher presents the principles of the European Union EU AI Ethics Guidelines 2026 (transparency, human responsibility, privacy, non-discrimination, well-being) and the groups reflect: “Do our rules cover these principles?”
• The rules are voted on and the Ethics Contract is posted in the classroom (to be revisited and updated in Lesson 6): the proposals are taken seriously, regardless of linguistic complexity.

Differentiation & Inclusion

Students with specific writing difficulties can do the concept map orally (audio recording) or with image support.

Part One – Introduction and engagement

At the start of this lesson, the teacher demonstrated the contrast between the results generated by an ineffective prompt versus an effective prompt in a generative AI tool. Students identify the differences between the prompts in terms of context, level, and specific request.

Ineffective prompt: “Tell me about solar energy”

Effective prompt: “What is the average efficiency of commercial photovoltaic panels in 2024 and how does it compare with the Shockley-Queisser theoretical maximum efficiency? Answer in simple terms for 9th-grade students.”

In a quick discussion, the class works together to improve both prompts (3 min).

Part Two – Exploration and application

For the research phase, the teacher assigns a topic and organizes the class into groups of 4. The topics are: Photovoltaic Solar Energy; Offshore Wind Energy; Hydroelectric Energy; Biomass and Biogas; Nuclear Energy; Fossil Fuels (oil and natural gas).

Within each group, students take on different roles: AI Researcher (formulates prompts), Verifier (confirms external sources), Data Analyst (handles quantitative data), Designer (infographic). Each group begins their work:

• Create prompts: each group formulates at least 3 progressive prompts, from general to specific, recorded in the Prompt Table of Worksheet Lesson 2.
• Test and evaluate the prompts: they test the prompts in a generative AI tool of their choice, record the AI’s summarized answer, and rate its quality on a scale of 1 to 5 (very poor to excellent).
• Information verification: each group must confirm at least 2 numerical figures generated by the AI (efficiency, installed capacity in Portugal, cost per MWh) against an official source. They record the URL and access date.
• Graphic synthesis (12 min): each group creates a slide or infographic with two sections: 1) information about their topic, energy source, operating principle (physical explanation), typical efficiency, advantages and disadvantages, estimated cost per MWh, use in Portugal (%); 2) “What the AI didn’t answer well or we had to verify”.
• Quick share (1 min): each group has 1 minute to present the source where it verified the information and what it questioned in the AI’s answer.

During the research, the teacher talks with each group and asks questions to help them assess the quality of the prompt and the AI-generated results:

Where does that efficiency value come from? Is it the same one you found on the IEA site? If it’s different, which do you trust more and why?

Differentiation & Inclusion

Groups with more difficulty receive a list of pre-written starter prompts; more advanced groups get the additional challenge of analyzing the source’s full life cycle.

Part One – Introduction and engagement

The teacher opens this lesson with an initial provocation that opens up space for the lesson’s central question:

If AI can be wrong and still seem credible, what do we do to protect the quality of our research?

The teacher projects 3 statements about energy: two correct and one generated by AI with a deliberate error. Students vote true or false by raising their hands. The surprise effect of the answers kicks off the discussion.

Part Two – Exploration and application

Activity 1: Error Hunt (20 min)

In pairs, students receive a worksheet with 6 AI-generated statements about energy, with deliberate errors of various types: wrong numerical figure; outdated information; incorrect generalization; concept confusion.

For each statement: they classify it (True / False / Don’t know), verify it against an external source, and correct or confirm it with justification.

Activity 2: Plenary sharing (8 min)

Students are invited to share: which errors were hardest to detect? Which type of error is most dangerous? Which verification strategies did they use?

Activity 3: EU Ethical Principles Game (12 min)

The teacher asks students to organize into pairs to play a game they will reflect on together at the end.

• Card Game (12 min): each pair receives one of the 5 ethical principles, which they must read, identify how it relates to what they did in the previous “Error Hunt” activity, and give a concrete example of how a user using a generative AI tool for research might violate or respect that principle.
• Presentation (1 minute per pair) + update of the Class Ethics Contract: the teacher connects the students’ contributions with the EU’s formal guidelines.

Activity 4: Individual Written Reflection (5 min)

At the end, the teacher asks each student to write an individual reflection on Worksheet Lesson 3, reflecting on:

“What did I change, or will I change, in the way I use AI after this lesson?”

Differentiation & Inclusion

Two-level worksheets: Level 1 with more obvious statements; Level 2 with subtle errors requiring prior knowledge (e.g., a wrong statement about thermodynamic efficiency). Students choose, or the teacher assigns by profile. A larger-font version for students with SEN; the option to work in a trio with peer support.

Part One – Introduction and engagement

In this lesson they explore what data is and how it is fundamental for training AI models like Machine Learning.

Activity 1

The lesson starts by introducing the concept of data, with the challenge of analyzing data from the Portuguese energy sector together, observing real-time data (5 min).

The teacher projects the REN dashboard in real time and asks students to observe and answer orally — recalling beforehand that AI is not a person, it can be wrong, it doesn’t know everything, and the information must be confirmed.

What do we see? What data is being shown? Who uses it? What is the point of having this data in real time?

Introduction to the concept of data (3 min)

Data is the raw material of AI — without training data there’s no Machine Learning. What does that mean for the energy sector?

Part Two – Exploration and application

To deepen understanding of the impact of data quality, the teacher shows the short video on AI in managing electrical grids (3 min) and facilitates a brief guided discussion:

What kind of data does AI use to predict consumption? What happens if the data is poor quality or biased?

Guided by Worksheet Lesson 4, the teacher presents the simplified scheme of the Machine Learning process (Training data → Model → Prediction → Evaluation → Adjustment) in Section B and asks the class to identify each phase in the concrete example.

• Real data analysis (15 min): each group of 4 students works with Worksheet A4: charts of the Portuguese energy sector 2015–2024 (energy mix composition, growth of renewables, consumption by sector, CO2 emissions). Guiding questions: What trends do you identify? In which sectors is consumption highest? Does the rise in renewables coincide with a reduction in emissions? What conclusions can you draw?
• Sharing: each group shares just one main conclusion with the class (5 min).

The AI and energy paradox (10 min): students use the green-algorithms.org calculator to estimate the carbon footprint of a typical ChatGPT interaction and of a large-scale AI model.

The AI we use to research energy… also uses energy. How do we balance that? What responsibility do we have as users?

Link to EU guidelines: the principle of sustainability and well-being, AI should be used in a proportionate and conscious way.

Differentiation & Inclusion

Worksheet Lesson 4 has charts at two levels: Level 1 with step-by-step guided questions; Level 2 with open questions requiring correlation between variables.

Part One – Introduction and engagement

The teacher opens the lesson by explaining the project activity students will develop. The challenge shared with students was:

You are consultants specialized in energy and sustainability. Your local Parish Council has received funding for a renewable-energy initiative and has asked for your technical proposal. You have 30 minutes to prepare it. Tomorrow is the Public Presentation.

The Project Guide is handed out to students (Worksheet Lesson 5)

Part Two – Exploration and application

Students organize into groups of 4 with defined roles (rotating relative to previous lessons): Solution Architect, Writer (AI + editing), Impact Analyst, Ethics Officer (ensures the grid is filled in). The stages of this activity are:

• Creation with AI (30 min): each group uses generative AI tools, ChatGPT or another, to generate text drafts for each Section; Canva or another, to create infographics and generate images; they integrate the real data verified in previous lessons. The group’s non-negotiable rule: every AI-generated paragraph must be re-read, discussed and edited by the group: the AI proposes, the group decides.
• Parallel completion of the Prompt Table in Worksheet A5: for each prompt used, they record what the AI generated, what the group changed and why.
• AI Use Declaration Grid (10 min): each group completes the full grid (what was AI-generated / what the group changed / what was decided exclusively by the group / external sources used to verify). This grid is an integral part of the final product.
• The teacher circulates among the groups and asks ethical and technical challenge questions.
• 2-minute preview per group: they present the title and central idea of the proposal to build anticipation for Lesson 6.

“Why did you choose this source of information? What data convinces you? Who could be harmed by this solution? Did the AI consider the stakeholders? What do you add?”

Differentiation & Inclusion

More advanced groups include a simplified cost-benefit analysis (estimated investment vs savings over 10 years); groups with more difficulty focus on the basic required elements.

Part One – Introduction and engagement

The lesson begins with students doing the final preparation for presenting the work their groups developed in the previous lesson. Groups are given 3 minutes to review the Presentation and make sure the “How we used AI and what we decided” section is clear.

The teacher recalls the assessment criteria and hands out Worksheet Lesson 6 (Peer Assessment Grid) to each student.

Activity 1

• Group presentations (27 min = 5 groups x 5 min + 2 min of questions): each group presents its work, mandatorily including: the energy source, the data supporting it, the stakeholder analysis, and the AI-use declaration section. The teacher may ask 1 ethical challenge question per group.
• Peer assessment: during each group’s presentation, every student in the class fills in the Peer Assessment Grid.
• Collective reflection: the teacher creates the “What I Learned Wall” (10 min), filled with one sticky note per student, answering one of three questions of their choice: “What did I learn about energy?”; “What did I learn about AI?”; “What will I do differently when I use AI in the future?”. The notes are posted on a wall in 3 columns. The teacher reads some aloud and synthesizes the unit’s learnings.

Part Two

• Re-reading the Ethics Contract (5 min): the Ethics Contract created in Lesson 1 is re-read in plenary. Going through each point one by one, students take a quick show-of-hands vote: do we keep it? Do we add something based on what we learned?
• Updating the Ethics Contract (5 min): the document is updated and approved by everyone.

Adelina shares a quote from the EU AI Guidelines 2026 and poses two final questions for anyone who wishes to answer:

“What is the role of the human in a world with AI? Did this unit help you answer this question?”

Differentiation & Inclusion

Students with oral-communication difficulties can present a more visual component or be supported by a classmate, what counts is the substantive contribution.

Adelina Silva has been teaching in public schools in Portugal for 37 years and has been doing research at Universidade Aberta in Portugal, for 24 years.

• Teacher of Technology Education in lower secondary and secondary school;
• Trainer certified by the Scientific Council for the Continuing Education of Teachers in the areas of: Administration; Computing; Secretarial studies; Educational Technologies (computing / applied computing);
• PhD in Anthropology from Universidade Aberta;
• Master’s in Intercultural Relations from Universidade Aberta.

The teacher’s work can be followed on her LinkedIn, Universidade Aberta Academia and ResearchGate pages.