4 min readMar 5, 2026 06:04 AM IST
First published on: Mar 5, 2026 at 06:04 AM IST
It would be dishonest to pretend that AI has not already changed how we write. For many people, it has made writing less intimidating. It helps organise scattered thoughts and improves language when words refuse to cooperate. For someone staring at a blank screen after a long day, it can feel like relief. A Fortune India report published last year found that 96 per cent of Indian professionals now use AI tools at work, including for routine writing tasks. But that convenience has not come without consequences.
The Supreme Court recently expressed disapproval of lawyers relying on AI tools to file submissions that cited non-existent or fabricated judgments. What took place in the courtroom is quietly unfolding in public writing. At Bharat Mandapam in New Delhi, leaders spoke eloquently of AI’s future. Outside those halls, AI is already editing our sentences, suggesting arguments, and sometimes drafting entire passages. Even so, readers are rarely told when a machine has played a meaningful role in shaping what they read. That omission matters.
Writing rests on a contract between the author and the reader. The latter offers attention, and the former offers a voice shaped by experience and judgment. When AI performs a substantial portion of the labour without disclosure, that contract weakens. The reader may not always detect it, but they sense a distance. The words are fluent, but the presence behind them feels thin.
Tools have always shaped writing. Dictionaries, editors and spellcheckers were once novelties too. The problem is not the tool but overreliance and silence about its use. When that silence becomes standard, the reader has no way of knowing whether they are engaging with a human mind or an autocomplete engine.
Access to advanced AI is not universal. When AI-assisted writing is consumed as wholly human work, comparison becomes unfair. A writer in a small town working without subscription tools is judged against prose refined by large language models they cannot afford. Disclosure does not eliminate this gap, but it can end the pretence that the field is level. More troubling is uniformity. As more writers lean on the same dominant systems, trained largely on Western language corpora, writing risks losing local texture and rhythm. Arguments fall into familiar grooves. In a country as linguistically diverse as India, such standardisation amounts to cultural loss. Over time, it can narrow the range of voices and perspectives that readers encounter.
What then is the remedy? Disclosure is often dismissed as performative or impractical. It need not be. Publications can ask writers to indicate the extent of AI involvement. If AI helped with language polishing, say so. If it drafted portions of the text, acknowledge that. The reader does not need technical detail, but they deserve clarity about whether the voice they engage with is entirely human. The law has already moved in this direction. Amendments to the Information Technology Rules in 2026 now require social media platforms to disclose AI-generated content.
None of these calls for Luddism or nostalgia. AI will remain part of how ideas are drafted and refined. The question is whether writers choose honesty over convenience and whether readers are given the courtesy of knowing how words were put together. In an age where machines assist expression, that small act of openness may be what keeps trust intact.
The writer is a lawyer who writes on public policy
