BS BFSI Summit: 'AI-powered devices, voice redefining digital banking'

AI is turning everyday payment devices into intelligent fintech tools, redefining banking, fraud detection, and customer interaction, experts said at the Business Standard BFSI Insight Summit

Devices that were once passive instruments in digital payments are now emerging as the new frontier of fintech innovation, as artificial intelligence (AI) reshapes the way banking, authentication, and customer interactions are conducted, industry leaders said at the Business Standard BFSI Insight Summit in Mumbai on Thursday. 

The panel discussion titled ‘From pocket to cloud: AI, authentication, and the future of digital banking devices’, moderated by Nivedita Mookerji of Business Standard, had views from experts from Google, Paytm, Zoho Payment Technologies, and Azentio Software. They discussed how AI is turning devices into intelligent computing systems that power real-time interactions and fraud detection. 

“Devices that used to be on the passive node are now actively engaged in the payments system. With AI, the way devices are evolving is huge,” said Ripunjai Gaur, chief operating officer (COO), offline payments, Paytm. 

He added that AI-enabled devices are making multilingual communication possible for merchants. 

“We have launched an AI soundbox, which can solve problems for merchants. With the AI device, a merchant can talk to anyone, for example, tourists,” Gaur said. 

Sivaramakrishnan Iswaran, chief executive officer (CEO), Zoho Payment Technologies, noted that banking has moved “from desktop to the cloud” and that AI is driving the next phase of contextual and voice-based banking. “Once it is on the cloud, it is all about AI. Voice is going to be the primary form of interaction in these devices,” he said. 

Mayank Sharma, regional head of Android Enterprise Partnerships at Google, said India has been a key player in AI-driven device innovation. “When Gemini Live was launched, India was one of the select markets. It was launched in nine Indian languages,” Sharma said, adding that much of such technology is “made in India by Indian engineers.” 

He also highlighted the evolution of smartphones and how they have become the centre of AI computation, and are backing the digital banking infrastructure. 

“The first Android phone came out in 2008. Since then, smartphones have changed from a connectivity delivery mechanism to full high power computing mechanism. We are working with a lot of banking institutions to leverage this AI technology on these devices,” Sharma added. 

On being asked about the rising number of fraud cases being reported and how these companies tackle it, Gaur said the industry has moved from manual to real-time fraud detection. 

“With computing capability increasing tremendously, we have a real-time fraud check mechanism. Today, you have UPI device banding. It restricts UPI transactions in case of theft, hacking, etc. However, companies still have to work on data security,” Gaur said. 

Iswaran said AI engines were improving the correlation of fraudulent patterns, which helps in fraud detection. 

However, he added, it increases the cost of operations due to which companies can’t reduce merchant discount rate (MDR).

“Once we all work collectively at the device level, technology level, and banking level, and raise red flags based on various parameters and empower the top layer, we can bring down fraud and that would lead to massive savings, which could be passed on to the customer,” he said. 

L Guru Raghavendran, senior vice president and product head at Azentio Software, said technology must serve the original purpose of banking while ensuring trust. 

“Purpose remains the same. But now it happens on a device and technology should support it,” he said, adding that devices now have “450+ parameters to check the authenticity of a transaction and whether it is a fraud". 

Looking ahead, Gaur said the future lies in building digital trust. 

“From icon-based browsing to voice-based browsing, applications will evolve. Then the cost can also come down, as we won’t need a screen. Evolving from digital reach to digital trust is important. The next 300-400 million people would transact digitally when they feel secure. We are working on how we can build trust,” he said.