A new study shows that essays drafted by ChatGPT read smoothly, yet feel distant. AI essays lack the subtle devices that pull readers in.
This gap may help teachers separate genuine coursework from machine writing. The finding urges stronger critical literacy in an age of rapid text generation.
Can AI write better than a student?
Researchers at the University of East Anglia, with a collaborator from Jilin University, worked on the comparison. They judged 145 student essays and the same number created by ChatGPT, scanning each for cues that invite a reader to engage.
The scientists hope the results will aid teachers and exam boards worldwide in flagging suspicious assignments. Software detectors still generate false positives, so human judgment remains vital.
By learning the discourse habits most students use, markers can notice when a script drops the personal edge that usually colors undergraduate prose. This knowledge supports fair grading and helps preserve academic integrity in classrooms already flooded with generative tools.
Students and the rise of AI
Professor Ken Hyland of UEA’s School of Education and Lifelong Learning voiced the worry driving the project.
“Since its public release, ChatGPT has created considerable anxiety among teachers worried that students will use it to write their assignments,” said Hyland, who warned of wider harm.
“The fear is that ChatGPT and other AI writing tools potentially facilitate cheating and may weaken core literacy and critical thinking skills. This is especially the case as we don’t yet have tools to reliably detect AI-created texts.”
AI writing is less engaging
The researchers searched for questions, personal asides, direct appeals, and other “engagement markers.”
“We were particularly interested in looking at what we called ‘engagement markers’ like questions and personal commentary,” Hyland said.
“We found that the essays written by real students consistently featured a rich array of engagement strategies, making them more interactive and persuasive.”
Human writers sprinkled their work with moments that addressed the reader, building a sense of dialogue.
“They were full of rhetorical questions, personal asides, and direct appeals to the reader – all techniques that enhance clarity, connection, and produce a strong argument.”
AI lacks the personal touch
“The ChatGPT essays, on the other hand, while linguistically fluent, were more impersonal. The AI essays mimicked academic writing conventions but they were unable to inject text with a personal touch or to demonstrate a clear stance,” Hayland explained.
Without questions or vivid commentary, the machine output felt flat. “They tended to avoid questions and limited personal commentary. Overall, they were less engaging, less persuasive, and there was no strong perspective on a topic.”
“This reflects the nature of its training data and statistical learning methods, which prioritize coherence over conversational nuance,” noted Hyland.
Classroom uses, not shortcuts
Even so, the authors do not reject AI tools. They see them as potential tutors when used openly.
“When students come to school, college or university, we’re not just teaching them how to write, we’re teaching them how to think – and that’s something no algorithm can replicate,” said Hyland.
The researchers urge teachers to design process‑based tasks that require drafts and reflection – steps no chatbot can provide. Training students to spot engagement markers, they add, will sharpen both writing and detection skills.
The study lands as detection software races to catch up with generative models. Commercial tools still stumble on hybrid texts that mix human and machine sentences.
For now, stylistic clues give teachers a slim but useful edge, though the target may shift as AI absorbs more conversational cues.
Preserving honest student voices
Coursework must remain proof of independent thought. If essays lose that role to AI, qualification systems wobble.
By showing where AI still falters, the study offers data to shape new safeguards while keeping space for honest, imaginative student voices.
The authors place the project within a wider push for digital literacy. Students meet machine text in news feeds, search results, and chat apps long before they step into lecture halls.
Teaching them to ask who wrote a sentence and why it was written is now central to education. Writing with a visible stance trains that habit. It also makes learners less willing to accept anonymous prose at face value.
The study is published in the journal Written Communication.
—–
Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.
Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.
—–