One Syllabus Bans AI. The Next Says Nothing At All.

analysis of texts · Frontiers in Education · la publicación, 19 ago 2026 · gratis

En 35 programas de curso, la IA se prohibió en 11 y no se mencionó en otros 11

Un estudio leyó 35 programas de curso de una universidad pública de Estados Unidos. Once no decían nada sobre el uso de IA, y eso no prueba que allí todo esté permitido.

Versión breve · la versión detallada sigue, unos 6 min

Pregunte al chat de weeklyAI

Pregúnteme por este estudio: a quiénes se estudió, qué encontró y qué no dice.

Las conversaciones se guardan mientras exista weeklyAI, para mejorar la publicación. Se responde en el idioma en que usted escribe.

El estudio, de un vistazo
Quiénes: programas de curso (syllabi) de una universidad pública de Estados Unidos
Cuántos: 35 programas de curso
Dónde: una universidad pública grande de Estados Unidos
Cuándo: cursos usados entre 2023 y 2026
Tipo de estudio: análisis de documentos (programas de curso)
Quién lo hizo: Universidad Estatal de Pensilvania (Penn State), Estados Unidos
El límite que importa: Son 35 programas de una sola universidad, sin criterio estadístico y con un solo analista.

Programas silenciosos y prohibitivos en Educación frente a otras áreas

Educación (22 programas)frente aOtras áreas (13 programas)

Silenciosos: siete en Educación frente a cuatro fuera de Educación

Educación (22 programas)frente aOtras áreas (13 programas)

Prohibitivos: siete en Educación frente a cuatro fuera de Educación

Un programa que calla es un hueco en el papel, no un permiso.
Lectura de weeklyAI

Así podría verse · ilustración generada por weeklyAI.watch, no es una fotografía

Una investigadora de la Universidad Estatal de Pensilvania, en Estados Unidos, juntó 35 programas de curso, esos documentos que los profesores reparten la primera semana. Veintidós eran de la facultad de Educación. Los otros trece venían de sociología, ciencias políticas, informática, ingeniería, física, estadística y otras carreras. Los cursos se dictaron entre 2023 y 2026. En cada documento buscó una sola cosa: qué le dice el curso al estudiante sobre usar inteligencia artificial para hacer sus tareas.

La universidad no había impuesto una regla única. Dejó que cada profesor decidiera si la IA se usa en su curso y cómo. La universidad pidió cinco cosas: que cada profesor decida si la IA se usa en su curso y cómo, que cada curso comunique sus reglas con claridad, que el uso permitido se declare y se cite, que el estudiante siga siendo responsable de que su trabajo sea correcto y honesto, y que el uso de IA respete la privacidad y los datos personales.

Lo que los programas decían era disperso. Once no mencionaban la IA para nada. Once la prohibían. Cinco permitían sólo usos acotados, como lluvia de ideas, esquemas o corrección de gramática. Dos la permitían en general, con condiciones. Y seis enseñaban la IA como tema del curso.

El motivo más repetido no fue el plagio. En 18 de los 24 programas que hablaban del tema, la razón era que el trabajo entregado fuera del propio estudiante. Por eso muchos dejaban pasar la ayuda pequeña y reservaban para el alumno el argumento, la interpretación y la escritura de fondo.

El estudio leyó documentos, no aulas. No puede decir qué explicó cada profesor en voz alta, ni en un correo, ni en las instrucciones de cada trabajo. Tampoco sabe si alguien fue sancionado. Un programa que calla es un hueco en el papel, no un permiso.

La universidad no tiene nombre en el artículo, y son 35 cursos de un solo país. Esto no describe las reglas de su universidad ni las de su país. Lo que lo obliga a usted es lo que digan su propia institución y los documentos de su propio curso.

Al empezar su próximo curso, busque el punto donde el programa habla de la IA. Si no hay ninguno, llévelo como pregunta al profesor antes de suponer que puede usarla.

Qué significa para usted

Lo que usted encontrará al empezar su próximo curso depende de su propia universidad, no de este estudio hecho en un solo país. Cuando llegue el programa, busque el punto donde hable de la inteligencia artificial. Si no dice nada, pregúntele al profesor antes de suponer que puede usarla, porque el silencio en el papel no es un permiso.

Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876

Quién pagó: La autora declaró que no recibió apoyo financiero para este trabajo ni para su publicación, y declaró no tener relaciones comerciales o financieras que pudieran constituir un conflicto de interés.

Versión detalladaLos pasajes copiados del artículo, las ilustraciones y cada fuente con cuánto leímos de ella · 1122 palabras · unos 6 minLeerla →Cerrar

Lo que dicen los programas de 35 cursos sobre el uso de la IA: casi un tercio no dice nada

Un estudio de una universidad pública de Estados Unidos revisó los programas de curso como textos que fijan reglas. En 22 de los 35, casi dos de cada tres, el programa prohibía la IA o guardaba silencio.

En una universidad pública de Estados Unidos, una investigación revisó 35 programas de curso usados entre 2023 y 2026 y encontró que las reglas sobre el uso de inteligencia artificial por parte de los estudiantes cambian radicalmente de un curso a otro. Once programas no decían nada sobre el tema, once la prohibían, cinco permitían usos concretos, dos la permitían en general con condiciones de responsabilidad, y seis trataban la IA o el aprendizaje automático como objeto de estudio profesional¹. Los dos grupos más grandes fueron, entonces, el silencio y la prohibición, con once programas cada uno².

Los programas analizados salieron de una universidad pública grande: 22 de la facultad de Educación y 13 de otras áreas. La institución no imponía una sola regla: dejaba la decisión en manos de cada docente, siempre que se comunicara con claridad, se exigiera atribución, se verificara la información y el estudiante siguiera siendo responsable de su trabajo³. Al comparar cada programa con ese marco, seis lo citaban de forma explícita, siete coincidían sin citarlo, ocho lo desarrollaban con detalle propio, tres daban apenas una mención y once guardaban silencio; ninguno contradecía directamente un requisito concreto de la institución⁴.

Aquí conviene entender qué es un programa de curso y por qué importa. No es solo una lista de lecturas: es el documento donde se comunican las expectativas, se definen las formas de ayuda aceptables y se fijan las condiciones bajo las cuales se evaluará el trabajo del estudiante, traduciendo principios generales en reglas concretas⁵. La universidad resuelve así un problema que su propia guía general no puede resolver: cómo definir el uso aceptable de IA dentro de un curso o una evaluación determinada⁶.

La razón más repetida para fijar reglas fue que el trabajo presentado fuera auténtico o producido de manera independiente: apareció en 18 de los 24 programas que sí decían algo sobre IA⁷.

El detalle vale la pena, porque muestra que la discusión no es solo sobre copiar. Una prohibición podía ser tajante: un programa de Educación decía que los estudiantes "no pueden usar herramientas de IA generativa para ayudar a completar el trabajo del curso"⁸. Otro, de ciencia de materiales, incluía el uso de herramientas como ChatGPT dentro del plagio y advertía que podía llevar a reprobar, aunque permitía ayuda limitada de edición⁹. En el extremo opuesto, un curso de estadística de Educación fomentaba el uso de IA para generar ideas, código de ejemplo, investigación y autoevaluación, siempre que apoyara los objetivos del curso y se reconociera su uso¹⁰. Y un curso sobre temas críticos de IA era, a la vez, un curso que estudiaba la IA y un curso que la restringía: se podía usar para lluvia de ideas, esquemas y gramática, pero había que documentarlo y no se podía entregar contenido sustantivo generado por IA¹¹.

El hallazgo central es que esta variedad no sigue una división limpia entre facultades. Las proporciones de programas silenciosos y prohibitivos fueron casi idénticas en los dos grupos: siete de 22 en Educación eran silenciosos, frente a cuatro de 13 fuera de Educación; y siete en Educación eran prohibitivos, frente a cuatro fuera¹². Entre los casos prohibitivos de otras áreas había cursos de desarrollo humano, métodos cualitativos, sociología y ciencia de materiales¹³. Los patrones de gobierno cruzaron las fronteras disciplinares, y todos los casos en que la IA era objeto de aprendizaje estaban en cursos directamente relacionados con ella¹⁴.

Una advertencia importante sobre el silencio: significa solo que el programa no documenta ninguna regla sobre IA. No significa que el docente no tenga posición ni que no haya dado instrucciones por otro medio, como una discusión en clase, una instrucción de tarea, un correo o la plataforma del curso¹⁵. El estudio tampoco midió qué pasó después: no dice si alguien fue sancionado, ni cómo se aplicaron las reglas, ni cómo las entendieron los estudiantes. Son 35 programas de una sola universidad, elegidos sin criterio estadístico, y el análisis lo hizo una sola persona sin un segundo evaluador independiente.

Así lo leemos nosotros. Cuando una autoridad reparte su poder y no fija un mínimo común, es fácil que las decisiones difíciles queden sin dueño visible: nadie firmó la regla, así que nadie responde por ella. En los cursos que no dicen nada, lo más probable no es que no exista un criterio, sino que se decidió en otro lugar sin quedar por escrito ni ser consultable. Si eso es cierto, el estudiante queda sin saber a quién obedece ni por qué. Sabríamos que nos equivocamos si resultara que esos cursos silenciosos son justamente los que más explican sus criterios en otros documentos accesibles al estudiante. Lo que usted puede hacer con esto es concreto: pida por escrito el criterio de uso de IA cuando el programa no lo diga, y guarde esa respuesta. Si usted enseña o decide, deje constancia escrita de la regla y de su razón, porque una regla no escrita no se puede reclamar ni discutir.

Y hay una segunda forma de mirarlo. Cuando el criterio descansa en la preparación y el buen juicio de una persona más que en una norma explícita, quien lo ejerce puede actuar con seguridad sin deber explicaciones, y quien lo recibe queda dependiendo de la buena voluntad de esa persona. En cursos donde el docente decide solo, la experiencia del estudiante variará según a quién le toque: unos explicarán con detalle por qué permiten o prohíben, y otros no darán ninguna razón. Nos equivocaríamos si los cursos con más libertad docente resultaran ser también los que más justifican por escrito sus decisiones. Por eso, al elegir cursos o al reclamar, fíjese menos en si la IA está permitida y más en si la razón está escrita y es coherente con lo que se le evalúa. Puede pedir esa razón; no es un favor, es lo que hace revisable la decisión.

La propia investigación apunta hacia dónde podría ir esto. Las universidades podrían exigir a cada docente que diga si la IA está permitida, prohibida o depende de la tarea; que identifique qué funciones se permiten y cuáles no; que aclare cómo se declara y se cita su uso; y que indique dónde buscar más orientación¹⁶.

Lo que este trabajo le deja a usted no es una regla, sino una pregunta útil para llevar al aula o a la coordinación: ¿qué criterio escrito se aplicó a mi trabajo, y quién lo firmó?

De dónde sale cada dato de contexto, y cuánto leímos de cada documento

Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Course-level enactment was highly heterogeneous: 11 syllabi were silent on student AI use, 11 were prohibitive, five permitted specified uses, two broadly permitted AI with responsibility safeguards, and six treated AI or machine learning as an object of pedagogical or professional learning."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Silent and Prohibitive were the two largest categories, each accounting for 11 of the 35 syllabi. A syllabus was coded as Silent when it contained no explicit guidance concerning student AI use."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The institutional framework delegated substantial authority to instructors while emphasizing communication, attribution, verification, and student responsibility."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Six syllabi were explicitly aligned with institutional guidance, seven implicitly aligned, eight elaborated the institutional framework, three provided minimal guidance, and 11 remained silent; none directly contradicted a specific institutional requirement."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "This study shifts attention from institution-level guidance to the course syllabus as a formal site of AI governance. Syllabi do more than describe course content and assignments. They communicate expectations, define acceptable forms of assistance, establish the conditions under which student work will be evaluated, and translate broader institutional principles into course-level rules."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "As generative artificial intelligence becomes embedded in digital education, universities face a governance problem that institutional guidance alone cannot resolve: how should acceptable AI use be defined within particular courses and assessments?"
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Authentic or independently produced work was the most common rationale, appearing in 18 of 24 non-silent syllabi."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "One Education syllabus stated that students “may not use generative AI tools to assist with the completion of coursework.”"
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "A non-Education materials-science syllabus stated that plagiarism included the use of generative AI tools such as ChatGPT and warned that violations could result in failure of the course, although limited editing assistance was permitted."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "One Education statistics syllabus encouraged AI use for idea generation, exemplar code, research, and self-assessment, provided that the use supported course learning outcomes and was acknowledged."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "One Education course on critical AI issues was Pedagogical in purpose but Conditional in its student-use rules: students could use AI for brainstorming, outlining, and grammar support but were required to document its use and could not submit AI- generated substantive content."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The proportions of Silent and Prohibitive syllabi were nearly identical across the two groups. Seven of 22 Education syllabi were Silent, compared with four of 13 non-Education syllabi. Seven Education syllabi and four non-Education syllabi were Prohibitive."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The non-Education Prohibitive cases included courses in human development, qualitative methods, sociology, and materials science. These cases demonstrate that restrictive governance was neither unique to Education nor limited to writing-intensive professional programs."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Governance patterns crossed disciplinary boundaries, while all pedagogical cases were concentrated in AI-adjacent courses."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Silence refers only to the absence of documented AI guidance in the syllabus. It does not establish that the instructor had no position on AI or that no guidance was provided through oral discussion, assignment instructions, email, or the learning-management system."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Universities could require instructors to state whether AI is permitted, prohibited, or assignment-dependent; identify relevant permitted and prohibited functions; clarify disclosure and attribution requirements; and indicate where students should seek further guidance."

Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876

analysis of texts · Frontiers in Education · the paper, 19 Aug 2026 · free

One Syllabus Bans AI. The Next Says Nothing At All.

A researcher read 35 course syllabi at one large U.S. public university. Eleven never mentioned AI, and what the other 24 said differed from course to course.

Short version · the longer version follows, about 6 min

Ask the weeklyAI chat

Ask me about this study: who was studied, what it found, and what it does not say.

Conversations are saved for as long as weeklyAI exists, to improve the publication. Answers come in the language you write in.

The study at a glance
Who: course syllabi at one large U.S. public research university
How many: 35 syllabi
Where: United States
When: 2023 to 2026
Kind of study: analysis of what people did
Who did it: Department of Education Policy Studies, Pennsylvania State University
The limit that matters: It read syllabi only, from one university, so it cannot show what was said or enforced.

Silent and prohibitive AI rules in Education versus other fields

Education syllabiagainstNon-Education syllabi

Seven of 22 Education syllabi were silent, compared with four of 13 non-Education syllabi.

Education syllabiagainstNon-Education syllabi

Seven Education syllabi and four non-Education syllabi were prohibitive.

A silent syllabus is a blank page in the document, not proof that anything goes.
weeklyAI's reading

How it could look · illustration generated by weeklyAI.watch, not a photograph

A syllabus is a formal document that communicates course obligations, assessment expectations, grading standards, and acceptable student conduct. On generative AI, at one large public research university in the United States, it often said nothing.

Evelyn Wu, a researcher in education policy at Pennsylvania State University, read 35 of these course documents. Twenty-two came from the College of Education. The other 13 came from sociology, political science, engineering, physics, statistics and other fields. All were used between 2023 and 2026.

The university itself had no single rule. It told instructors to decide for their own courses, while asking that expectations be spelled out, that permitted AI use be disclosed and credited, and that students stay responsible for the accuracy of their work.

What the syllabi did with that freedom was scattered. Eleven said nothing about AI at all. Eleven banned it. Five allowed only named uses, such as brainstorming or checking grammar. Two welcomed it broadly, with safeguards. Six taught AI or machine learning as the subject of the course.

The most common reason given was not cheating. In 18 of the 24 syllabi that mentioned AI, the aim was authentic work: the paper, the analysis, the argument should be the student's own thinking. Many courses allowed small help, like an outline or a grammar check, and kept the argument itself for the student.

The study has a gap it cannot close. It read syllabi only. It cannot show what instructors said out loud, what they wrote in emails, what students understood, or whether any rule was ever enforced. A silent syllabus is a blank page in the document, not proof that anything goes.

It also comes from one university in one country, so it describes a pattern, not your rights. The rules that bind you are the ones your own institution and your own course documents state.

When your next course starts, look for where the syllabus states its AI rule. If there is none, that absence is a question to put to the instructor.

What this means for you

When your own course starts, find the line in the syllabus that says whether AI is allowed, and if there is no such line, ask the instructor directly, because a blank page is not permission. The study says nothing about your country, your school or your rights, so treat what it found as one university's habit and not as a rule that binds anyone.

Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876

Who paid: The author declared that no financial support was received for this work or its publication, and declared no commercial or financial relationships that could be a conflict of interest.

The longer versionThe passages copied from the paper, the pictures, and every source with how much of it we read · 1124 words · about 6 minRead it →Close

Your Course Says Nothing About AI. That Silence Has a Rule of Its Own.

A study of 35 syllabi at one U.S. public university found course-level AI rules ranged from outright bans to open encouragement—and nearly a third said nothing at all.

If you are a student, the rule that actually binds you on using AI for an assignment is probably not your university's. It is your course's. A new study examined 35 syllabi from a large public research university in the United States, used between 2023 and 2026, and compared what each one said about generative AI with the institution's own guidance. The corpus was 22 syllabi from the College of Education and 13 from other fields.¹²

The finding was not one policy but many. Eleven syllabi said nothing at all about student AI use. Eleven prohibited it. Five permitted specified uses, two broadly permitted AI with responsibility safeguards, and six treated AI or machine learning as the object of study itself. Silent and prohibitive were the two largest groups, tied at 11 each.³⁴

The institution had not failed to write a policy. It had written one that handed the decision to instructors, while asking them to emphasize communication, attribution, verification and student responsibility.⁵ Measured against that framework, six syllabi explicitly echoed institutional guidance, seven were implicitly compatible, eight went further and spelled out distinctions the institution left general, three offered only minimal guidance, and 11 were silent. None directly contradicted a specific institutional requirement.⁶

The technology here is the ordinary kind: text generators that will draft, summarize, rewrite or answer a question on request. The governance question is narrower and harder: in a given course, is using one of those tools assistance or substitution? A syllabus does not settle that by describing the tool. It settles it by naming which parts of the work belong to the student.²

The reasons instructors gave matter more than the bans. Across the 24 syllabi that said something, the most common rationale was authentic or independently produced work, appearing in 18 of them—more common than academic integrity, which appeared in 12.⁷ That is a different worry. Integrity asks whether a rule was broken. This asks whether the submitted work still shows what the student knows.

The patterns did not sort by faculty. Silent and prohibitive policies appeared at nearly identical rates in Education and outside it: seven of 22 Education syllabi were silent against four of 13 elsewhere; seven Education syllabi and four others were prohibitive. The prohibitive cases outside Education included human development, qualitative methods, sociology and materials science.⁸⁹ Every one of the six syllabi that taught AI as a subject was in a course where AI or machine learning was central to the curriculum—two inside Education, four outside.¹⁰

Concrete examples show how far apart two courses at the same university can be. One Education syllabus stated plainly that students "may not use generative AI tools to assist with the completion of coursework."¹¹ A materials-science syllabus counted generative AI tools such as ChatGPT as plagiarism and warned that violations could mean failing the course, while allowing limited editing help.¹² At the other end, an Education statistics course encouraged AI for idea generation, example code, research and self-assessment, provided the use served the learning goals and was acknowledged.¹³ And one Education course on critical AI issues allowed brainstorming, outlining and grammar support but barred AI-generated substantive content.¹⁴

What the study cannot tell you is what any of this meant in a classroom. It read documents. A silent syllabus does not mean the instructor had no position—expectations may have been given aloud, by email, in assignment instructions or on the course platform.¹⁵ The analysis was done by one researcher with no second coder, so the categories are transparent judgments rather than fixed measurements. It covered one institution and cannot describe universities generally.

Here is how we read it—and the first thing to say is that we are reading, not reporting. When a rule is left unwritten, the rule does not vanish; it gets carried by whatever the instructor does, how assignments are marked, and what classmates say is normal. So our expectation is that a silent syllabus will not feel neutral to you. It will feel like a risk to be guessed at, and the safest guess is usually the most cautious one. You would show us wrong by telling us your silent course felt open—that you simply asked and got a straight answer, or used your own judgment without worrying. If you are in a course that says nothing, ask in writing before the first assignment and keep the reply: what is allowed, what must be disclosed, and who to ask if the answer changes.

One more thing we would watch for. Where the syllabus is silent, two students in the same room can end up following two different rules—one from something the instructor said in passing, one from a classmate's assumption—and neither is written anywhere either of them could point to. That is not a prediction about your course. It is a reason to notice what your instructor actually does with AI in their own materials and slides, and to ask your classmates what they were told. If the rule only exists in someone's head, ask for it on paper before it matters.

We also think it is worth knowing, before you need it, who at your institution decides an academic-integrity case involving AI, and what the steps are if you disagree. The study does not tell us how any of that worked in practice—it read syllabi, not case files. But a rule you cannot appeal is a rule you cannot argue with, and the time to find out whether there is a person at the end of the process is not the week a grade is held up. Ask now; keep the answer with the one you got from your instructor.

What would have to change is modest and specific. Universities could require instructors to state whether AI is permitted, prohibited, or depends on the assignment; name which uses are in and which are out; spell out disclosure and attribution; and say where students should go for more guidance.¹⁶ The study also notes that privacy and confidentiality—concerns the institution itself raised—never appeared as a stated reason in any of the syllabi it read.

So the useful move is not to wait for a policy to arrive. Read your own syllabus for what it actually says, and treat the blank space as something to fill by asking rather than by guessing. Then ask yourself the question this study leaves standing: in the work you are about to hand in, which parts are meant to be yours—and do you know, in writing, where the line sits?

Where each piece of context comes from, and how much of it we read

Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "As generative artificial intelligence becomes embedded in digital education, universities face a governance problem that institutional guidance alone cannot resolve: how should acceptable AI use be defined within particular courses and assessments?"
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "This study shifts attention from institution-level guidance to the course syllabus as a formal site of AI governance. Syllabi do more than describe course content and assignments. They communicate expectations, define acceptable forms of assistance, establish the conditions under which student work will be evaluated, and translate broader institutional principles into course-level rules."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "Course-level enactment was highly heterogeneous: 11 syllabi were silent on student AI use, 11 were prohibitive, five permitted specified uses, two broadly permitted AI with responsibility safeguards, and six treated AI or machine learning as an object of pedagogical or professional learning."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "Silent and Prohibitive were the two largest categories, each accounting for 11 of the 35 syllabi. A syllabus was coded as Silent when it contained no explicit guidance concerning student AI use."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "The institutional framework delegated substantial authority to instructors while emphasizing communication, attribution, verification, and student responsibility."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "Six syllabi were explicitly aligned with institutional guidance, seven implicitly aligned, eight elaborated the institutional framework, three provided minimal guidance, and 11 remained silent; none directly contradicted a specific institutional requirement."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "Authentic or independently produced work was the most common rationale, appearing in 18 of 24 non-silent syllabi."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "The proportions of Silent and Prohibitive syllabi were nearly identical across the two groups. Seven of 22 Education syllabi were Silent, compared with four of 13 non-Education syllabi. Seven Education syllabi and four non-Education syllabi were Prohibitive."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "The non-Education Prohibitive cases included courses in human development, qualitative methods, sociology, and materials science. These cases demonstrate that restrictive governance was neither unique to Education nor limited to writing-intensive professional programs."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "Governance patterns crossed disciplinary boundaries, while all pedagogical cases were concentrated in AI-adjacent courses."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "One Education syllabus stated that students “may not use generative AI tools to assist with the completion of coursework.”"
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "A non-Education materials-science syllabus stated that plagiarism included the use of generative AI tools such as ChatGPT and warned that violations could result in failure of the course, although limited editing assistance was permitted."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "One Education statistics syllabus encouraged AI use for idea generation, exemplar code, research, and self-assessment, provided that the use supported course learning outcomes and was acknowledged."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "One Education course on critical AI issues was Pedagogical in purpose but Conditional in its student-use rules: students could use AI for brainstorming, outlining, and grammar support but were required to document its use and could not submit AI- generated substantive content."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "Silence refers only to the absence of documented AI guidance in the syllabus. It does not establish that the instructor had no position on AI or that no guidance was provided through oral discussion, assignment instructions, email, or the learning-management system."
Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876 - the article this story is about — the whole article — the passage: "Universities could require instructors to state whether AI is permitted, prohibited, or assignment-dependent; identify relevant permitted and prohibited functions; clarify disclosure and attribution requirements; and indicate where students should seek further guidance."

Wu, E. (2026). Governing generative AI in digital education: how institutional guidance becomes course-level policy. Frontiers in Education. https://doi.org/10.3389/feduc.2026.1886876

Who paid: The author declared that no financial support was received for this work or its publication, and declared no commercial or financial relationships that could be a conflict of interest.