Como universidades estão criando regras para o uso de inteligência artificial
Área principal: Education
Áreas secundárias: Universities | Science & Technology
Nível: B1
Como universidades estão criando regras para o uso de inteligência artificial
How Universities Are Creating Rules for AI Use
Proibir tudo ou permitir tudo já não descreve bem o debate universitário sobre inteligência artificial. Em 2026, instituições e órgãos educacionais estão desenvolvendo políticas mais detalhadas sobre quando, como e por que estudantes podem utilizar IA.
Contexto
A inteligência artificial generativa tornou-se uma questão de política acadêmica. Universidades precisam decidir como tratar ferramentas capazes de escrever, programar, resumir, traduzir e analisar informações.
Em julho de 2026, a OCDE identificou cinco áreas emergentes nas políticas de ensino superior: orientação para uso responsável, regras para conformidade e aquisição de ferramentas, desenvolvimento de competências em IA, avaliação por meio de pesquisas e pilotos e desenvolvimento de ferramentas especializadas. Ao mesmo tempo, universidades como Harvard e Stanford adotam modelos nos quais regras institucionais coexistem com instruções específicas para cada disciplina ou atividade.
READING TEXT IN ENGLISH
How Universities Are Writing the Rules for AI
A student opens an assignment and asks an important question:
“Can I use AI for this?”
The answer may be:
Yes.
No.
Or:
Yes, but only in specific ways.
This illustrates one of the challenges universities face in 2026.
Artificial intelligence can help students generate ideas, receive feedback, analyse information and practise skills.
But it can also complete work that students are expected to do themselves.
For universities, the challenge is creating rules that protect learning and academic integrity without ignoring a technology that is becoming part of professional and academic life.
The OECD reported in July 2026 that governments and higher-education systems are increasingly developing policies for responsible generative AI use. Important concerns include data protection, academic integrity, equity, reliability and the effect of AI on student learning.
Different universities are developing different approaches.
At Harvard's Faculty of Arts and Sciences, instructors are required to tell students what the generative AI policy is for their courses. A professor may restrict AI, allow particular uses or design activities that deliberately include it.
Stanford follows a similar course-level approach. Its guidance recommends that instructors clearly identify whether AI use is prohibited, limited, assigned or open, and communicate the rules in the course syllabus.
This means that the same AI activity may be acceptable in one course but unacceptable in another.
Why?
Because learning objectives are different.
In a writing course, producing the first draft may be an important part of learning.
In another course, AI may be intentionally used to analyse and critique information.
Rules also involve privacy.
Universities warn students and staff not to upload sensitive or protected information to unapproved AI systems. Harvard, for example, warns that information entered into some public AI tools may not remain private.
The emerging principle is therefore not simply:
“Use AI”
or
“Don't use AI.”
It is:
Know what you are allowed to use, why you are using it, what you must disclose and which intellectual work must remain yours.
Key Idea
As políticas universitárias estão se tornando mais específicas porque AI use não é uma única atividade.
Existe grande diferença entre:
pedir definição de um conceito;
receber feedback sobre um texto;
gerar ideias;
corrigir gramática;
produzir todo o trabalho;
enviar material confidencial;
utilizar IA durante uma prova.
Harvard e Stanford, por exemplo, dão espaço significativo para que professores definam regras adequadas aos objetivos de suas disciplinas. A OCDE observa, por sua vez, que instituições e governos precisam trabalhar também com privacidade, integridade, competências em IA e avaliação dos impactos educacionais.
A tendência é sair da regra genérica e caminhar para uma pergunta pedagógica:
What type of AI use supports the learning objective of this activity?
General & Academic Vocabulary
1. Policy
Significado: política / regra institucional
Definition: An official set of principles or rules.
Collocation: university policy
Example: The university introduced a new AI policy.
2. Academic integrity
Significado: integridade acadêmica
Definition: Acting honestly and responsibly in academic work.
Example: AI policies must protect academic integrity.
3. Permitted
Significado: permitido
Definition: Officially allowed.
Example: AI use is permitted for this activity.
4. Prohibited
Significado: proibido
Definition: Officially not allowed.
Example: AI is prohibited during the examination.
5. Disclosure
Significado: declaração / divulgação
Definition: The act of making information known.
Example: The professor requires disclosure of AI use.
6. Guideline
Significado: diretriz
Definition: Advice or rules explaining how something should be done.
Example: Students should read the AI guidelines carefully.
7. Syllabus
Significado: programa/plano da disciplina
Definition: A document describing a course and its requirements.
Example: The AI policy appears in the course syllabus.
8. Data privacy
Significado: privacidade de dados
Definition: Protection of personal or confidential information.
Example: Data privacy is an important concern when using AI.
9. Assignment
Significado: trabalho / atividade acadêmica
Definition: A piece of work students must complete.
Example: Students completed the assignment independently.
10. Learning objective
Significado: objetivo de aprendizagem
Definition: Something students are expected to learn or be able to do.
Example: AI use should support the learning objective.
11. Responsible use
Significado: uso responsável
Definition: Use that follows ethical, academic and practical rules.
Example: Universities are teaching responsible AI use.
12. Compliance
Significado: conformidade
Definition: The act of following rules or requirements.
Example: Institutions need clear processes for AI compliance
Language Focus
Permission and Obligation
Políticas acadêmicas utilizam frequentemente estruturas para dizer o que alguém:
can docannot domust doshould do
CAN
Students can use AI to brainstorm ideas in this assignment.
= é permitido.
CANNOT
Students cannot use AI during the exam.
= não é permitido.
MUST
Students must disclose their use of AI.
= é obrigatório.
SHOULD
Students should check AI-generated information carefully.
= é recomendado.
Observe como mudar apenas o modal altera completamente a regra.
Check Your Understanding
Why is it difficult for universities to create one simple rule about AI?
What concerns does the OECD identify around generative AI in higher education?
Who can establish specific AI rules for courses at Harvard and Stanford?
Why might AI use be acceptable in one assignment but not another?
Why is data privacy part of university AI policy?
Critical Thinking
Should every university use exactly the same AI rules? Why or why not?
Should students always disclose when they use generative AI in academic work? Explain your position.
Mini Academic Task
Write an AI Policy
Imagine que você é professor universitário.
Crie uma política de 120–160 palavras para sua disciplina.
Inclua quatro partes:
AI is permitted for...
AI is not permitted for...
Students must...
Students should...
Use pelo menos uma vez:
can — cannot — must — should — however — because
Termine explicando:
The purpose of this policy is...
Answer Guide
1. Because courses and assignments have different learning objectives and AI can be used in many different ways.
2. Data protection, academic integrity, equity, reliability of AI-generated information and possible effects on student knowledge and skills.
3. Individual instructors can establish course-specific rules within broader institutional guidance.
4. Because one activity may use AI as a learning tool while another may be designed to assess the student's independent ability.
5. Because sensitive, confidential or protected information can be exposed when entered into some AI systems.
Sources
Policies Supporting Responsible and Systematic GenAI Adoption in Higher Education — OECD Education Spotlights — July 17, 2026.
Generative AI Guidance — Harvard Office of Undergraduate Education — current guidance requiring faculty to communicate course policies for generative AI use.
AI Course Policies — Stanford Center for Teaching and Learning — guidance distinguishing prohibited, limited, assigned and open AI use.
Guidelines for Using ChatGPT and Other Generative AI Tools at Harvard — Harvard Office of the Provost — guidance on privacy, accuracy and academic integrity.
Guidance for Generative AI in Education and Research — UNESCO — global human-centred policy framework, updated January 2026.
Conheça LevX General & Academics e desenvolva o inglês necessário para estudar, pesquisar e comunicar-se em ambientes internacionais.
Comentários