Written by Diogo Matos, Systems Analyst at FCT, in the digital services unit FCCN
Artificial Intelligence (AI) has gone from being a novelty to becoming part of our daily lives. We often use it without realizing it, in different everyday tasks. Education is no different: it can help teachers save time, support students in their learning, and contribute to more efficient processes. But, like any technology with a growing impact on our lives, it requires clear criteria, critical thinking, and responsible use.
Throughout this article, we share a set of best practices and essential principles for using AI in an informed, responsible, and conscious way, maximizing its benefits without losing sight of the associated challenges and risks.
Topics in this article:
1. Transparency above all
The responsible use of Artificial Intelligence begins with transparency. Whenever we use these tools to support or produce a work, a text, an analysis, or a decision, we must clearly disclose this fact. This applies to students who use this technology for schoolwork, to teachers who use it to prepare materials, and to any professional who integrates it into their work process.
Transparency is not just an ethical issue; it allows those who receive the work to correctly evaluate what they are reading and how it was produced.
2. A tool, not a replacement.
Artificial intelligence models can be valuable allies in various tasks, from content creation and information analysis to organizing ideas, performing repetitive tasks, or identifying aspects that may have been overlooked. However, their role should be supportive, not a replacement.
The responsibility for the final decision, whether it's a grade assigned, a diagnosis, a report, or a simple email reply, always remains with a person. This means that the result produced by AI should always be critically analyzed, rather than automatically accepted. Delegating judgment entirely to an AI model is to abdicate our own responsibility.
3. Fact-checking is a mandatory step.
Models can be wrong, and they are more often than you might think. So-called "hallucinations," where incorrect information is presented with complete confidence, are a real and known risk. Dates, names, citations, bibliographic references, or statistical data generated by Artificial Intelligence should always be verified with reliable sources before being used or shared.
A good practice is to ask the model to indicate the references and sources used to support the information presented. These references should then be consulted and verified directly by the user to confirm their accuracy, reliability, and suitability to the context.
4. Data security and privacy
Many of the available tools collect user data for analysis, service improvement, and, in some cases, training future models. Therefore, it is essential to avoid entering personal, confidential, or sensitive data on these platforms, whether it belongs to us, our students, or our colleagues.
Before using a document or asking a question that involves identifiable information, it's worth stopping and thinking: "Could this content leave this space and be used for other purposes?" When in doubt, it's preferable to anonymize it or simply not share it.
5. Pay attention to biases.
All models are trained on large datasets, mostly collected from the internet and other existing sources. This means they may reflect and sometimes amplify biases present in that data, such as gender stereotypes, cultural prejudices, geographical imbalances, or other forms of unequal representation of reality.
Users should be wary of answers that appear biased, stereotypical, or incomplete, and should always question the origin and neutrality of the information received.
For teachers: transforming principles into practice.
In education, the responsible integration of Artificial Intelligence involves not only adopting these tools, but also the ability to guide students towards a critical and conscious approach. In this context, there are some practices that can make a difference inside and outside the classroom.
Define and communicate clear usage policies.
Responsible use begins with clear rules. Students must understand when and for what purposes Artificial Intelligence can be used in academic work, as well as the limits of its use. For example, it may be suitable for supporting research or organizing ideas, but not for writing an entire final paper.
Digital literacy: understanding how AI works
More than just using the tools, it's important to understand how they work and their limitations. Students should know that these systems can make mistakes, present incorrect information, or reflect biases present in the data they were trained on. They should also understand that the quality of the answers depends, to a large extent, on how the questions are formulated.
Reduce the administrative burden.
Artificial Intelligence can be a valuable ally in reducing the administrative workload of teachers, freeing up time for what really matters: monitoring students and supporting their learning process. This technology can contribute to tasks such as lesson planning, structuring curricular units, or preparing summaries and syntheses of the content taught.
These tools can also facilitate the creation of exercises, tests, and worksheets, the adaptation of materials to different levels of difficulty and specific educational needs, the development of assessment rubrics, and the correction of assignments. In all these cases, review and validation by the teacher remain essential.
Leading by example in the use of AI.
When a teacher uses Artificial Intelligence in a transparent and critical way, they are also promoting these good practices among students. Clearly identifying when it has been used, verifying the information produced, and maintaining professional judgment as a central element of the decision are ways to demonstrate, in practice, what responsible use means.
Preparing for the future with a critical mind.
Artificial Intelligence has been transforming the way we study, teach, and work. When used judiciously, it can support learning, simplify tasks, and increase efficiency in different contexts. However, its potential must go hand in hand with transparency, critical thinking, and respect for privacy. More than knowing how to use these tools, it is essential to understand their limitations, validate the information they produce, and always keep human judgment at the center of decisions.
In a context where AI is increasingly present in our daily lives, developing an informed and responsible use of it is an essential skill for everyone. Technology evolves rapidly. The responsibility for using it appropriately remains entirely ours.
