Telcos mull deploying AI to check pesky calls amidst surge in fraud

Mobile phone operators are exploring deploying artificial intelligence (AI)-led interventions to address the issue of pesky calls, with some saying they may even extend the blockchain-based solution currently being used for messages to calling.

“Some telcos have started tinkering with the idea of using AI/ML (machine learning) for addressing the issue of pesky calls, but it is at very nascent stages,” said an industry executive, who did not wish to be identified.

Emails sent to Reliance Jio, Bharti Airtel and Vodafone Idea were unanswered till press time.


The move comes even as mobile users are being increasingly bombarded by the menace of pesky calls and messages, leading to thousands of crores of financial fraud. A whopping 66% of mobile users continue to get at least three pesky calls every day, most of which originate from personal mobile numbers, according to a national survey by LocalCircles.

The AI-based technology involved will be able to recognise spam call patterns and alert the network operators which can then take appropriate action, said industry executives and experts.

“A technology intervention that studies call patterns like duration of calls and the frequency of outgoing calls is a solution that telcos will need to deploy. And this will have to be across telcos for it to be effective,” said Faisal Kawoosa, chief analyst, Techarc.

The numbers used for spam calls usually have high outgoing traffic, with each call lasting for a short duration. Discerning such patterns is possible by deploying AI/ML-led solutions.

Extending the distribution ledger technology (DLT) concept to calls is the obvious next step, according to the executives. The DLT-based verification process, in addition to AI/ML-led spam call detection, could effectively help reduce pesky calls in the long run, they said.

DLT is a blockchain-based solution that the Telecom Regulatory Authority of India (Trai) has mandated all operators use to keep spam messages in check.

Essentially, legitimate marketers and telemarketing agencies need to register themselves, the headers (the code that one sees when receiving messages from a brand) and the message templates on the DLT platform. Only registered senders with valid headers and matching templates will be allowed to send messages under this regime.

“All the registered enterprises will have a legitimate calling ID in text form. The CLI (caller line identification) of the legitimate calls from the enterprises will have a text-based call ID,” said Jyotirmoy Chakravorty, founder of Ubona Technologies, a provider of cloud telephony.

CLI refers to the number that flashes on a receiver’s screen when it is not saved on their contact list.

Trai has aggressively taken up the issue of pesky or spam calls, also called unsolicited commercial communication. It recently released a consultation paper on mandatory implementation of ‘calling name presentation’ in telecommunication networks which will display the name of the caller on the receiver’s screen when the call connects, even if the receiver has not saved it.

It has also commissioned a pilot by Vodafone Idea and Tanla Platforms of an AI-based solution to curb phishing and cyber frauds emanating from misuse of SMSes.

The regulator recently announced the formation of a joint committee of regulators comprising the Trai, Reserve Bank of India, Securities and Exchange Board of India and the consumer affairs ministry to address the issue of pesky calls and messages.