Why voice is becoming India’s next payment frontier
India’s Unified Payments Interface has made mobile payments ubiquitous in the subcontinent. As the country gears up for voice-activated transactions, experts warn of new risks involving AI and audio deepfakes
In less than a decade, India and the Unified Payments Interface (UPI) have become synonymous. With even vegetable sellers processing payments using a smartphone, the UPI system is a technological marvel that many countries now view as a benchmark for digital payments.
Since its inception in 2016, the number of UPI transactions has increased from 20 million to more than 241 billion. Transaction values rose from 70 billion rupees ($730m) to about 314 trillion rupees ($3.3tn) in under a decade, while the number of participating banks grew from 21 in April 2016 to more than 700 over the same period.
By July 2025, the UPI system – touted as the world’s number one real-time payment network, surpassing Visa – was serving 491 million individuals and 65 million merchants. It accounted for 85% of all digital transactions in India and 50% of such transactions globally. UPI has also been supported by countries such as the United Arab Emirates, Singapore, Bhutan, Nepal, Sri Lanka, France, Mauritius and Qatar for cross-border payments.
But alongside UPI’s widespread adoption, the challenges have compounded. QR code phishing and other payment frauds, such as PIN scams and impostor requests, have been worrying threats. According to a 2025 report by PwC, UPI fraud losses in 2022-23 totalled about 5.73bn rupees ($60m), accounting for 0.4% of total transaction value.
A year later, that figure grew to nearly 10.9 billion rupees ($113m), or 0.5% of the transaction value. Estimates show 1.34 million incidents of domestic UPI fraud in the 2023-24 fiscal year, and another 632,000 incidents involving 4.85 billion rupees ($50.5m) through September of the 2024-25 fiscal year.
Now, engineers are asking: What if voice can be used to perform UPI transactions? Could that make real-time payments more secure, faster and accessible to more users, particularly those lacking digital literacy, modern devices or reliable internet connectivity?
The advantage of sound over touch
“Voice UPI refers to enabling UPI payments through voice commands instead of traditional app navigation or typing,” said Mehul Mistry, senior vice-president of customer success, growth and strategy for India at Zeta, a banking technology company.
“In a Voice UPI flow, users can simply speak instructions such as, ‘Send 500 rupees to Rahul’, ‘Pay electricity bill’, or ‘Transfer money to this shopkeeper’,” said Mistry. “The system leverages AI [artificial intelligence], speech recognition and UPI infrastructure to understand the request, authenticate the user and complete the transaction seamlessly.”
Voice UPI has the potential to become the next major evolution of the platform because it expands digital payments beyond screens and literacy barriers, said Vivek Kumar Singh, co-founder and director of labs at ToneTag, a deep tech company building voice-first payment technology.
As conversational AI, multilingual voice interfaces and embedded fintech solutions continue to evolve, Voice UPI is likely to become a mainstream financial inclusion layer that makes digital payments more natural, trusted and universally accessible
Vivek Kumar Singh, ToneTag
“The next phase of inclusion will come from intuitive, conversational and assisted commerce experiences, especially for small merchants, rural users, senior citizens and first-time digital adopters,” said Singh.
Pratim Mukherjee, senior engineering director at McAfee, noted that voice-enabled payments have the potential to make digital transactions easier and more accessible, particularly for people who may be more comfortable speaking than typing or navigating apps.
India has already seen early versions of Voice UPI emerge across the ecosystem, with conversational payment assistants and voice-enabled payment journeys being explored by banks and financial technology companies, Mistry added.
One major breakthrough expected from voice-based payments is a reduction in QR-related fraud.
“QR-based transactions still face issues such as fake or tampered QR codes, merchant confusion around payment confirmation, and dependency on visual interfaces,” Singh said, adding that Voice UPI can add a more intuitive and trusted layer through real-time audio confirmations, guided prompts and conversational assistance. “Another major opportunity is in low-connectivity environments, especially in semi-urban and rural regions where internet reliability remains inconsistent.”
A key advantage of Voice UPI is its flexibility across multiple channels and devices, spanning smartphones, banking apps, AI assistants, chatbots, call-based systems, smart speakers, wearable devices and basic feature phones. Mistry noted that the technology has also spawned use cases in areas such as highway toll payments, insurance premium collections, utility bill payments and e-commerce payments at the point of delivery.
Voice-enabled payments also offer significant benefits for senior citizens, vernacular-language users, first-time smartphone users, the visually impaired, and consumers in semi-urban and rural markets who may be uncomfortable navigating complex app interfaces, said Ramkumar Subburaj, co-founder and chief technology officer of Phi Commerce, a digital payments firm.
Security challenges
In populous markets such as India, a spoken payment command can be overheard, recorded and replayed.
“That’s not innovation; that’s an attack surface,” warned Aaron Bugal, field chief information security officer for the Asia-Pacific region at Sophos. “Add AI voice cloning, where seconds of audio are enough to impersonate someone, and fraud stops looking suspicious and starts sounding normal.”
Mukherjee is also wary of the security implications. “As payments become something we can authorise with our voice, scammers will likely look for ways to exploit that trust,” he said. “AI is already making it easier to clone voices.”
He cited McAfee research that found 70% of people globally were not confident they could tell the difference between a cloned voice and a real one. “As this technology continues to evolve, safeguards like identity verification, transaction confirmations and consumer awareness will be important to help people use it with confidence,” said Mukherjee.
Even when funds are not stolen, damage is done. “Voice payments leak metadata: who you pay, how much and when,” said Bugal. “That data fuels social engineering, account takeovers and identity fraud – all for the sake of a little extra convenience.”
To mitigate such security threats, Subburaj said voice transactions will need layered security models that leverage voice recognition, device binding, behavioural analytics, transaction risk-scoring, PIN verification and AI-driven anomaly detection.
“Voice AI systems must also recognise regional accents and dialects, as well as handle noisy environments and multilingual conversations,” he added. “It is not only a payment challenge but also a matter of AI maturity.”
The future of voice payments
Nonetheless, Singh is confident that ecosystem readiness for Voice UPI is stronger today than ever before.
“India already has the foundational infrastructure in place, including high UPI penetration, widespread merchant acceptance and growing AI adoption,” he said. “As conversational AI, multilingual voice interfaces and embedded fintech solutions continue to evolve, Voice UPI is likely to become a mainstream financial inclusion layer that makes digital payments more natural, trusted and universally accessible.”
Subburaj concurred, noting that interoperability, Aadhaar-based digital identity, inexpensive data access, high smartphone penetration and an innovation-driven ecosystem established by the National Payments Corporation of India are laying a solid foundation for voice-led payments.
“Merchant readiness is also elevated, as most businesses now have QR-enabled acceptance infrastructure,” he added. “Voice UPI will likely grow into a complementary payment channel in the short term, rather than a replacement for current UPI interfaces.”
Mistry noted that the biggest opportunity for Voice UPI lies in driving financial inclusion through conversational and multilingual payment experiences. “Combined with AI and lightweight infrastructure, Voice UPI can simplify digital payments for millions of users beyond urban smartphone-native consumers, particularly in vernacular and low-literacy markets,” he said.
While open source AI gained unprecedented recognition during the India AI summit, divisions over governance, market concentration and regulatory power cast doubt on whether the technology will benefit society as a whole.
OpenAI is preparing a one-gigawatt datacentre in India as part of its $500bn Stargate infrastructure push, in what would be its biggest bet yet on its second-largest user base.
