Essay writing is one of the most controversial applications of generative artificial intelligence (GenAI) in higher education. Models can produce well-structured texts in a matter of seconds using a single prompt. In fact, recent versions of tools such as ChatGPT demonstrate writing performance that surpasses undergraduate students’ abilities. This reality forces educators to confront a central question: should students use GenAI to write essays, and if so, under what conditions? Rather than adopting a simplistic stance of unconditional acceptance or outright rejection, I suggest a nuanced and pedagogically grounded approach based on four recommendations.
The first is straightforward: use GenAI. At first glance, this may seem unnecessary or even provocative, especially given that a significant sector of the educational community views GenAI as a threat to academic integrity and intellectual development. However, refusing to acknowledge its presence is both unrealistic and counterproductive. Students already use it regularly, not only for academic purposes such as summarising texts, proofreading or drafting ideas, but also for everyday tasks including cooking, exercise planning or seeking advice. In this context, guided and supervised use by faculty is far preferable to indiscriminate, unsupervised and potentially harmful use.
Crucially, such guidance is impossible if instructors themselves lack familiarity with these tools. Faculty members must understand how these models work, what they can and cannot do and when their use is pedagogically appropriate. AI literacy among educators is, therefore, a prerequisite for meaningful regulation of student use.
The second recommendation is more restrictive: do not use GenAI at certain stages of the writing process. For instance, although GenAI excels at organising and synthesising information, excessive reliance on these functions may weaken students’ ability to retain information and develop mental mapping skills. These skills are not limited to essay writing; they are essential for problem-solving, critical thinking and lifelong learning. Structuring arguments, identifying relationships between ideas and organising complex information are fundamental intellectual competencies. For this reason, instructors must be explicit about when AI use undermines learning objectives and should therefore be avoided.
- Spotlight guide: Bringing GenAI into the university classroom
- Is critical thinking the answer to generative AI?
- Should we kill the essay?
The third recommendation follows naturally: establish a clear usage plan. To maximise the benefits of GenAI while minimising the risks associated with its use, instructors must specify precisely when students can and cannot use it. The implicit rule that “what is not prohibited is permitted” creates ambiguity and encourages misuse. Simply stating that “GenAI is prohibited in this course” is neither realistic nor effective. Instead, policies should articulate concrete guidelines, such as: “You may use GenAI at this stage, but you must not use it at that stage.” Determining these boundaries requires expert judgement and depends on the course’s learning objectives, reinforcing the importance of faculty GenAI literacy.
In my teaching practice, I recommend using GenAI to help define a main idea, generate an initial outline, or discuss potential arguments that students have already identified. In these cases, AI functions as a cognitive scaffold or supervisory assistant rather than a substitute for thinking. Conversely, I strongly discourage using GenAI to replace reading assigned materials, interpreting texts, generating original arguments, or writing the final version of an essay. In short, AI should accompany the writing process, not replace it.
The fourth recommendation addresses enforcement: when in doubt, interview without accusing. Even with clear guidelines, students may still violate GenAI usage rules. Importantly, such violations do not always stem from malicious intent. Poor time management, lack of confidence in writing skills, underestimating the importance of an assignment or a simple misunderstanding of the rules are common factors. While we should not ignore or excuse deliberate misconduct, instructors must exercise caution before assuming dishonesty.
In my experience, an effective approach is to ask students to explain how they wrote a particular section by narrating their process. I then ask them to explain the content itself, including key concepts and relationships. If inconsistencies emerge that suggest improper GenAI use, I ask the student to rewrite the passage in their own words, based on the explanation they just provided. This strategy often yields more coherent and authentic writing, even if penalties still apply.
The challenges GenAI poses to essay writing are significant and continually evolving. Nonetheless, outright rejection is neither feasible nor desirable. A balanced approach grounded in clear rules, pedagogical intent, and instructor competence offers a viable path forward. By defining when students should and should not use AI, and how potential violations are addressed, educators can transform GenAI from a perceived threat into a powerful educational ally.
AI use declaration: I declare that I used the OpenAI language model ChatGPT 5.2 to review the English translation and to refine the style. The final text, the critical analysis and the conclusions are the result of my own authorship and personal review.
Luis Moisés López Flores is an educator at Tecnológico de Monterrey, Mexico.
If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.
comment