Why not simply normalise the use of artificial intelligence in academic or any other forms of writing? Some will call this an ethical violation, arguing that human intellect cannot be surrendered to thoughtlessness, and that people will hand the tasks to AI and then claim the outcome as their own. But have we not already crossed that bar? Taking credit from our mechanised tools.

The potatoes a chef puts into an air fryer are sold as the “chef’s special”. The shooter gets the credit or blame for pulling the trigger, not the metallurgist or the arms manufacturer. The cosmonaut is remembered as the one who landed on the moon, and not the thousands of engineers and taxpayers who indirectly made that expedition possible. In every one of such cases, we already assign authorship to the person who directs the tool, not to the tool itself, or to every component that contributed to it. Yet, when the tool is generative AI, the same logic is suddenly reversed. The individual who directs it is accused of forfeiting authorship rather than exercising it. How come it is legitimate for a professor or researcher to claim authorship, despite taking all the help from their research assistants, such as collected and digitised data, field observations, case studies and narratives? Why such dualism?

It appears that this is not even a question of why, but whose. Whose authority is being protected by that resistance and challenged, if AI writing gets the green signal? With a good research idea and access to advanced AI tools, someone who has never set foot on remotest sites such as Antarctica or the Everest can now produce research that rivals the work of those who spent years in the field. Someone who never became bilingual can now write fluently across languages. Someone who always felt short to sharpen their own prose can now have it handed to them, polished, on a plate. Whose authority does that threaten?

Parallels in history

History gives a consistent answer as every novelty that redistributes access to power is resisted, and the resistance is rarely about the novelty itself. People resisted metal tools when stone tools were displaced. They resisted motorised wheels when animal carts were phased out. They resisted paper currency when it replaced gold and silver coins. The high-born and propertied classes resisted the idea of social reformers that everyone deserves a right to education. Every novelty has been resisted, from Galileo’s heliocentrism to the visibility of a non-binary identity in contemporary societies. And each time, the resistance came because the novelty threatened an existing arrangement of power.

Now, the question arises: who, then, benefits from labelling AI writing an “ethical violation”, and who stands at a disadvantage? Much of the loudest objection comes from a milieu long accustomed to a Western (Euro-American-centric) and elitist model of knowledge production. It is this claim of legitimate scholarship, a standard by which knowledge produced anywhere in the world is deemed worthy or unworthy. And AI has quietly punctured one of that milieu’s deepest advantages — the near-monopoly of native English fluency. A non-native speaker can now go toe-to-toe with a native English writer, with AI’s help. Outside the Anglophone world, standard “proper” English has been the single biggest challenge for brilliant minds, and AI has turned that barrier into an opportunity. In other words, a genuine redistribution of linguistic capital. The West has lost a linguistic monopoly it once did not have to defend. The discomfort around AI writing under the banner of “No, AI Please” tracks that redistribution more closely, tracking any coherent theory of authorship. AI writing has emerged as a genuine de-colonial tool for post-colonial societies, pushing back against the malpractices normalised under literary neocolonialism.

I wonder how long this charade can hold, as one can notice the selective normalisation. It is entirely acceptable to let AI generate references, citations, infographics, graphs, slide decks, and let it silently correct the grammar too. But letting it touch the sentences is not acceptable. This very selective normalisation has created an entire economy of tools such as “AI Content Humanizer”, “5th-Grade Style Writer”, “No-Fluff Humanizer”, “Natural Tone Rewriter”, “Conversational Rewriter”, and dozens more to launder AI prose back into a humanised pattern. It is strange that instead of letting AI help people cross language barriers and let their ideas flow, we normalise the malpractice of hiding it. In other words, the rule is not “don’t use AI”, rather, it is “don’t get caught”. As a professor of mine once put it, if you are going to lie, lie well enough not to get caught. And if you can’t, don’t lie at all. Consider J.K. Rowling’s Platform 9¾ at London’s King’s Cross Station, a lie told with such conviction that it convinced the world of a magical place that never existed but people now travel to see.

Not plagiarism

Now, before moving on to the arguments against AI writing, one must establish that AI use is not plagiarism. Plagiarism is theft of another’s ideas and work. AI-assisted writing is neither and entirely a separate debate. The objections raised against AI writing can be located into four broad strands: that it threatens originality and the meaning of scholarly work, heavy reliance on it dulls critical intellect, it risks leaking respondents’ personal data, and that it erodes human creativity. On reflection, scholarly worth has never been simply about originality. Illogical outcomes, rubbishness, and genocidal ideologies, too, have claimed “originality” within their own internal logic. What matters is contribution with a logical solution to the problem statement without harm. As for intellect, human habits of mind take generations to change, not possible within a span of a technology cycle. If anything, AI is capable of solving in minutes what once took months of manual labour, freeing attention for much more grave problems rather than dulling it. Third, data leakage is not an inherent property of tools such as AI, but a research-design failure, as pseudonymisation and anonymisation remain the researcher’s own ethical responsibility. And lastly, AI is not Alzheimer’s for human creativity. Better tools have never suppressed creativity. They have always redirected creativity.

Now, once these arguments are set aside, what remains is not a principled position but a familiar social reflex. The one that promotes “I have suffered through this, so you should too.” It reflects in professors’ posts on LinkedIn that they won’t read a draft touched by AI. Sounds similar to one of patriarchy’s oldest tricks: the mother-in-law who insists her daughter-in-law endure what she endured, or my office clerk who advised me to pile work on a new hire, stating why should they get an easy run when we did not. Schadenfreude dressed as rigour, a bizarre satisfaction taken in someone else’s hardship. Moreover, it is not a coincidence that the loudest objections tend to belong to those who have already cleared the language-proficiency bar the rest of the world is still trying to reach.

To conclude, AI writing is a redistribution of who gets these standards, rather than the erosion of standards. The only practical choice left is whether to accept that now, or to keep resisting a novelty to be accepted at a later stage, like everyone before it. It is inevitable that it will become the new normal, and failing to see that is not caution, but a failure of foresight. In the words of historian E.H. Carr, history can only be written by those who have a sense of direction and an instinct that they are advancing towards an end goal in order to piece evidence together convincingly. And man moves within history himself and indeed progresses.

Amit Kumar is an Assistant Professor in the Institute for Development and Communication, Chandigarh