For many international multilingual students in the United States, writing in English has always involved negotiation. We translate more than words. We translate experiences, cultural references, humor, emotion, and sometimes even versions of ourselves—into language that will be understood and accepted in an American classroom.
Then generative AI entered the picture.
Tools such as ChatGPT can make academic writing faster and less stressful. They can help a student brainstorm, clarify an idea, organize a paragraph, or find a more precise English expression. Yet beneath that convenience is a difficult question many multilingual students are quietly asking: If AI makes my writing sound more “native,” does it still sound like me?
Helpful, but Not Simple
The public conversation often divides students into two groups: those who use AI and those who do not. The reality is far more complicated.
In my research with multilingual student writers, I found that students’ attitudes toward generative AI were rarely entirely positive or negative. Many valued its support, particularly when writing in an additional language. AI could reduce anxiety, save time, and help them communicate ideas they already had but struggled to express in expected academic English.
At the same time, students were cautious. They worried about inaccurate information, academic-integrity rules, overdependence, and the possibility that polished AI-generated language might erase their individual voice. Their use of AI was not simply acceptance or rejection. It was a continuing process of deciding when the tool was helpful, when it became too influential, and where they needed to draw a boundary.
When “Better” Writing Feels Less Authentic
International students often arrive in U.S. classrooms already aware that their English may be judged against an imagined native-speaker standard. Their accents, sentence structures, and rhetorical choices can be treated as problems to correct rather than meaningful parts of their multilingual identities.
Generative AI can intensify that pressure. With a single prompt, it can transform a personal, culturally shaped piece of writing into smooth, standardized English. The revision may appear more professional, but it can also feel unfamiliar—too formal, too generic, or, as some students describe it, “too clean.”
This creates a strange dilemma. Students may feel that their own writing is not polished enough, while AI-assisted writing may be viewed as inauthentic or even suspicious. They are asked to improve their English but may be questioned when that improvement seems too sudden or too perfect.
In that space, multilingual students must protect two things at once: their academic credibility and their right to sound like themselves.
The Invisible Work of Negotiation
For these students, using AI responsibly involves much more than typing a prompt and copying a response. They compare the output with what they intended to say. They reject words they would never use. They restore examples that reflect their cultures and lived experiences. They adjust sentences until the writing feels recognizable again.
This is identity negotiation: deciding how much assistance to accept without allowing the technology to take ownership of the message.
It is also emotional labor. International students must interpret different professors’ expectations, unclear or changing AI policies, and the fear that any use of technology may be judged harshly. Because multilingual writers have historically faced greater scrutiny over language and authorship, the stakes can feel especially high.
What Students Are Really Asking For
Multilingual students are not asking for unlimited or uncritical AI use. They are asking for clearer guidance, fairer conversations, and space to make thoughtful choices.
Instead of treating generative AI only as a shortcut or a threat, educators can help students examine what a tool changes, what it removes, and whose version of “good writing” it reproduces. Students need opportunities to discuss acceptable uses, disclose assistance without fear, verify AI-generated information, and retain authority over their ideas and language.
Most importantly, they need reassurance that multilingualism is not a weakness that technology must hide.
Generative AI may help students find the words, but it should not decide who they become on the page. The central question is no longer simply whether international multilingual students use AI. It is whether they can use it while preserving the histories, identities, and voices that make their writing their own.
