At 9:17 in the morning and before the first customer walks in, 54-year-old fashion designer Neera Malik is on her phone.
She has just recorded a video in her boutique in Gurugram, showcasing a sleeveless blouse with detailed stone embroidery. It’s meant for her nearly 11,000 Instagram followers. She posts videos on every design she’s proud of—from bridal lehengas, waistcoats, to a dress that walked the runway in the Miss India contest.
But, as good as she is with her craft and handiwork, words elude her when describing her own creations. She takes to her phone again. She doesn’t type. She speaks.
A year ago, she would have waited for her son to come home or called her daughter for help with captions for her social media posts. Hashtags confused her. Trending music was a mystery. Now, she taps the microphone icon on the ChatGPT app and says, “Write a caption for a summer pastel suit. Modern, elegant, for working women. Add good hashtags.”
Within seconds, the response appears: “No sleeves, but no compromise on design.” The caption is in Hindi, complete with emojis, hooks, and music suggestions.
ChatGPT doesn’t always get it right. Sometimes, she has to prompt repeatedly before she gets what she wants. But she follows what her 25-year-old son, Harsh Malik, had once told her in jest: “Ask it anything. If it says something wrong, scold it. Make it correct itself.”
Today, the woman who once hesitated to press ‘Upload’ now posts daily. And she has even tasted fame, with a video peaking at 8.5 lakh views.
In many ways, her story mirrors her son’s.
AI for Everything
Harsh works as a software engineer at an investment bank in Bengaluru. He grew up in an Air Force family, where childhood was measured not in years but in cities.
“I’ve lived in 13 cities,” he says, smiling. “Changed schools six or seven times.”
That early exposure to adjusting to the new and unknown prepared him for the transition to artificial intelligence. He doesn’t fear it. He studies it. Uses it. Sometimes, even “misuses” it.
In the glass-and-steel corridors of his office, the bay of his blockchain team offers a quiet retreat, where he experiments with the technology.
“There are two parts to our job,” he tells AIM. “First is designing a system for a problem statement. Second is coding out that design. AI is helping a lot with the second part these days.”
“For unit test cases, I misuse AI,” he admits. “It is correct in most cases, but you have to guide it a lot. Otherwise, it does things which are not good practice.”
Of course, it is not perfect. At times, it overshoots its prompt or misinterprets intent.“Those kinds of things do happen, and sometimes we face problems,” he says.
His day-to-day work, however, is now more streamlined. Documentation that once took hours now comes together quickly. Small design diagrams—modelling a system’s structure, behaviour, and architecture—are easier to produce. Sequence diagrams, written in rigid languages where a single arrow requires multiple lines of syntax, no longer feel tedious.
Instead of wrestling with format, Harsh simply describes the interaction and the design. Within seconds, the UML (Unified Modelling Language) is generated.
Harsh is one of millions of developers in India who find themselves at an inflection point. They are increasingly building with AI at the centre, and not as an afterthought.
India accounts for 5.8% of Anthropic’s Claude model use, second only to the US. The AI tool boasts over 18.9 million monthly active users on its website. Harsh, too, prefers Claude Opus for code—and that reflects in the numbers, with India ranking first globally in the share of Claude’s use for software-related tasks.
Over 40% of software development work across Indian IT and tech companies is done by AI, according to NASSCOM.
But many Indian developers are now finding themselves heavily reliant on AI. In the pre-GenAI era as well, only a handful of Indian software engineers could write functionally correct code, and now more of them are dependent on it for sheer survival. Many, especially junior, inexperienced developers, can’t even recall the exact syntax for a simple for-loop.
Developers aren’t alone.
Writers use AI for drafts. Students depend on it for homework. X is full of posts asking Grok to process any new development. The phenomenon is known as cognitive offloading, when AI tools are used to store or process information, thereby conserving internal cognitive resources.
ChatGPT, Claude, Gemini, etc., have undoubtedly simplified daily lives and enhanced professional efficiency. But perhaps too simplified. There’s a subtle but certain undercurrent of overt GenAI use. It’s so pervasive that it now defines the normal.
According to Atlassian, 77% of Indian knowledge workers now use generative AI tools daily. Moreover, more Indians are using AI tools for everyday use than in any other country, per Statista Consumer Insights.
However, habitual reliance on AI tools may also risk cognitive inertia—a reduced tendency for reflective thinking and analytical reasoning. It risks giving rise to an AI-dependent generation.
Typing Away Everyday
AI has slipped into Harsh’s smaller decisions, too. He recently wanted to get a haircut. So, he uploaded his photo to ChatGPT, which he has subscribed to.
“I told it: I want something modern but not too sick. Smart, but not too much,” he recalls.
The model suggested styles that suited his facial structure, and even drafted instructions he could give his barber. “It also told me how to maintain it. How to style it. Everything,” he says, smiling. This constant reliance on the AI tool has shot up his screentime to 8 hours.
Sitting on the couch opposite him, his girlfriend, 26-year-old Anushka Aggarwal, tries not to giggle. “You could have just asked me,” she quips. She has watched him turn to AI for everything from workouts to weekend plans.
Anushka also works at the same investment bank, where they chase the same deadlines. Much like her boyfriend, she uses tools like GitHub Copilot to code, test, and document faster than she ever imagined.
“For coding, I use Copilot,” she explains. “When I join a new project, and there is a new code base, instead of going to people first, I take KT (Knowledge Transfer) through AI.”
AI spills over to her personal life as well. “Out of work, to design my outfits, jewellery, and footwear, I use ChatGPT,” she says.
She has subscribed to ChatGPT, too, which helps her communicate better. “If I am talking to someone professionally, to improve my conversation, I use it.”
Before sending a sensitive email or preparing for a difficult discussion, she refines her thoughts through AI. It has become her sounding board. But it has also pushed her screentime to 10 hours.
The 95% Developer
Across India, a new generation of developers now lives between chatbot prompts and midnight deploys. On most nights, 26-year-old Tanish Bansal also finds himself coding away.
Not the old way, when developers sat hunched over Stack Overflow tabs, stitching together snippets. Stack Overflow, too, is losing relevance as a community platform, with website traffic in steady decline since the launch of ChatGPT. That era, he says, is over.
“Two years ago, we used to copy from here and paste there,” Tanish narrates, leaning back in his chair. “Now? 95% of the code is written by AI. I just review it.”
Tanish grew up in Haridwar and studied computer science at Dehradun Institute of Technology. Today, he works remotely for Soulbound, a US-based Web3 startup. He primarily uses Claude, occasionally ChatGPT, and sometimes experiments with newer models.
AI, he believes, has flattened hierarchy. “Now juniors and seniors are on the same baseline. Everyone uses AI.”
An intern in his team produces code at a similar speed. The difference lies elsewhere. “Seniors know how things break. They know how to prompt better. They understand deployment risks,” he explains.
It reflects how AI is reshaping the Indian IT workforce. Companies are now abandoning the pyramid hierarchy, which was designed for effort-based billing and predictable delivery. The bottom is falling out, with entry-level roles being increasingly automated. The top five IT firms added only 17 net employees in the first nine months of FY26.
“We are not doing capacity-based pricing,” IT services company Mphasis’ CEO Nitin Rakesh had said during the Q3 FY26 earnings call. “Whether I need the same number of people, half the people or one-third of the people, the client has no consideration. What the client cares about is the outcome.”
At the same time, Gen Z employees want greater freedom to experiment with AI tools. According to a Canva survey, 78% of Gen Z workers in India desire experimentation even as organisational systems lag.
The Urge for Answers
The habits, attitudes, and ideas of Gen Z adults tell a far more fraught story.
OpenAI CEO Sam Altman had said last year that “older people use ChatGPT as a Google replacement”, whereas “people in their 20s and 30s use it like a life advisor.”
According to Harvard Business Review, roughly a third of young adults reported turning to AI for help with their personal life, one in four reported using chatbots “as a friend,” and one in ten said they used an AI chatbot “as a girlfriend or boyfriend.”
Tanish’s routine is a microcosm of that dependency. He treats ChatGPT almost like a second opinion. Over the past year, AI has slipped into almost every part of his day. Work. Health. Money. Decisions.
A regular gym-goer, he uses it to track his meals, asking how many calories a dish contains or whether he has hit his protein target for the day. When a health report showed elevated ESR levels, he consulted with the app about diet changes and additional blood tests. “It told me which test to choose. So I chose that. I didn’t overthink,” he relays.
That’s dangerous territory, but Tanish is among more than 40 million people who consult the platform daily for health information. However, many studies suggest that while some LLMs excel at identifying medical issues, they may mislead users in certain medical scenarios.
But Tanish goes beyond that. He has even vibe-coded an expense tracker.
“I wanted to track exactly where my money was going, so I built my own app. Whatever feature I need, I just add it. I don’t need to buy any product anymore.” He is convinced that SaaS products will die eventually.
Across the room, his girlfriend Aditi Chandra, a 27-year-old product designer at Tech9, stops using her smartphone. She has heard this before.
Aditi uses the same tools. She knows what they can do. But she doesn’t hand them this much of herself. Unlike Tanish, she is careful and thinks twice before uploading anything, whether it is client data, personal reports, or screenshots.
To her, AI is helpful. It is impressive. But it is still just a tool.
She uses AI for UX purposes, despite the hype that it would replace all the design jobs. Her company is pushing AI to become a part of the process. But she’s clear about its limits. It can help with speed and exploration. It can’t do all the creative work. “It is still a tool at this point,” she asserts.
Claude has become part of her design process. Wireframing no longer begins with crumpled sheets of paper.
“Earlier, we used to go to physical pen and paper, try layouts, how to arrange and all that. Now with Claude, it has become easier. I can say, give me six layouts of this, and it will try. Maybe four won’t be great, but at least you can brainstorm.”
She uses ChatGPT for generative experiments. Mock-ups. Visual explorations.
Her discomfort is not with AI as much as with the way people surrender to it. “People start trusting AI more. Which I feel is wrong,” she says.
There are inherent biases and hallucinations pervasive across AI systems trained on imperfect, and often, selective data.
A 2025 Stanford study found that LLMs and image generators systematically perpetuate stereotypes against older women, favouring older men. Meanwhile, cloud company Workday is facing a class-action lawsuit over using AI-powered applicant screening that unfairly discriminates against candidates aged 40 or older, as well as those with disabilities.
Aditi is measured when it comes to using AI outside work. “It’s not like I ask every single question throughout the day,” she says. There is a difference, she insists, between assistance and dependence.
Aditi and Anushka say they and their boyfriends are not losing sleep over AI, at least not yet. It’s more wait-and-watch, curious about what comes next.
“There’s no real worry right now,” Anushka says. “As long as you’re good at what you do and know how to use AI tools to your advantage, it feels manageable.”
They are aware that things are changing, and maybe faster than expected, but they haven’t reached a tipping point yet. “Harsh is not at a point in his career where he sees it affecting him enough to feel anxious,” she notes.
However, she admits there is a quiet unease about what the future might look like.
Aditi describes a similar mindset. “Tanish is not worried. It is more about understanding how to work with it instead of stressing over it. Right now, he is more excited about what can be built and what is possible.”
While there is a persistent uncertainty, it has not turned into fear. Right now, there is a sense of possibility and a willingness to see where things go.
Your reaction