 "Airtel")
With the solution, Airtel intercepts 100 million suspected spam calls and 3 million suspected SMSes from 2 million suspected connections daily (Photo: Shutterstock)
Bharti Airtel on Wednesday announced a network-based solution that will use artificial intelligence (AI) and 250 parameters to detect spam calls and messages for the telecom company’s customers.
The solution was developed in-house and starting Thursday it will use a proprietary algorithm to identify and classify calls and text messages as “suspected SPAM”, said Gopal Vittal, managing director and chief executive officer of Bharti Airtel. It has been shown to successfully identify 97 per cent of spam communication.
The solution analyses real time parameters such as a caller or sender’s usage patterns, call/SMS frequency, and call duration. By cross-referencing this information against known spam patterns, the system accurately flags suspected spam calls and SMSes.
The solution is based on 250 parameters, including the velocity and volume of calls, frequency of device changes, number of short-duration calls, geographical spread of calls, time on the network, number of names on the same KYC, and robo-calling devices, to analyse whether an incoming call or SMS is spam.
Airtel’s network size sets the solution apart from existing platforms such as Truecaller, according to Vittal. The solution has two filters: One at the network level and the other at information technology systems. Every call and SMS passes through this dual-layered “AI shield”, which has learned from Airtel’s network traffic over the past few months.
“In two milliseconds, our solution processes 1.5 billion messages and 2.5 billion calls every day. This is equivalent to processing 1 trillion records on a real-time basis using the power of AI. Our solution has successfully identified 100 million potential spam calls and 3 million spam SMSes originating every day,” said Vittal.
Leveraging technology
The solution will alert customers to malicious links received via SMS. To achieve this, Airtel has built a centralised database of blacklisted URLs, and every SMS is scanned in real-time by a state-of-the-art AI algorithm to caution users against accidentally clicking on suspicious links. However, Vittal emphasised that Airtel will not read the content of the messages.
The solution can also detect anomalies such as frequent IMEI changes, which are typical indicators of fraudulent behaviour. By layering these protective measures, the company ensures its customers receive maximum protection against the evolving landscape of spam and fraud threats.
With the solution, Airtel intercepts 100 million suspected spam calls and 3 million suspected SMSes from 2 million suspected connections daily.
However, the solution is still a work in progress, particularly in expanding its coverage to feature phones and video calls. WhatsApp calls will also not be covered by the solution, as they are fully encrypted.
First Published: Sep 25 2024 | 2:37 PM IST
Leave a Reply