Is Voice becoming the front door to AI for India’s next billion users?

The auto rickshaw was weaving through Delhi traffic with its usual disregard for physics when something else caught my attention.

At a red light, the driver did not glance at his phone or tap a screen. Instead, he spoke, clearly and confidently, into the air. “Hey Google, maps mein navigation shuru karo.” Instantly, Google Maps sprang to life, charting a course to my destination. No typing. No fumbling. Just voice.

It was not an isolated moment. Over the past year or so, I have seen kirana shop owners chatting away to AI chatbots, asking questions about stock, margins or even how to word a WhatsApp broadcast to customers.

On another occasion, while I was conversing with Gemini using voice, an acquaintance standing beside me watched in fascination before demanding a crash course on how to “talk” to ChatGPT and Google’s AI instead of typing to it.

These vignettes may feel anecdotal, but together they point to something bigger. Voice is quietly, and rapidly, becoming the front door to artificial intelligence in India.

For decades, digital adoption in India has been shaped by constraint. Small screens, patchy connectivity, multiple languages, varying literacy levels and a general impatience with keyboards. Voice neatly sidesteps most of these frictions. It is intuitive, immediate and deeply cultural. We Indians talk, a lot, and we talk in many languages, often switching mid sentence.

Big Tech has noticed.

OpenAI’s latest India focused ad film leans squarely into this reality. The protagonist is not a coder or a startup founder, but a kirana shop owner fretting over customer retention as a shiny new mall opens nearby.

His sounding board is an AI assistant, accessed through conversation rather than commands. Strategy, loyalty and competition, all classic MBA themes, delivered in a tone that feels more chai pe charcha than corporate offsite.

Google, meanwhile, has been even more explicit. Its recent Pixel campaigns position the device as an AI-first smartphone, with Gemini Live, the conversational, voice led interface of Gemini, taking centre stage. 

And the data backs this up. According to a 2024 Google blog, over 40% of Gemini’s Indian language users already rely on voice interactions. That is a staggering figure when viewed through a global lens.

Google has since rolled out Gemini Live in Hindi, followed by eight more Indian languages, Bengali, Gujarati, Kannada, Malayalam, Marathi, Telugu, Tamil and Urdu, doubling down on localisation as a growth lever.

This did not begin with generative AI. As far back as 2021, Google reported a 270% surge in voice searches in India. Voice notes on WhatsApp, voice search on YouTube, voice commands on Android. Indians have been rehearsing for a voice first AI future without realising it.

What generative AI does is elevate voice from utility to interface. No longer is it just about “navigate here” or “set an alarm”. Now, voice enables reasoning, creativity and problem solving, tasks that previously demanded literacy, digital fluency or both.

For the next wave of users, especially those coming online via affordable smartphones and prepaid data, speaking to an AI may feel less like using technology and more like consulting a knowledgeable assistant.

Of course, challenges remain. Accuracy across accents, trust, hallucinations, data costs and privacy concerns will all shape adoption curves. But the direction of travel feels unmistakable.

That auto rickshaw driver did not think he was “using AI”. He was simply talking to his phone and getting things done. And perhaps that is the point. When technology disappears into habit, adoption stops being a headline and starts being infrastructure.