Airtel’s system works by passing every call and SMS through two filters — one at the network level and another at the IT systems layer. Airtel claims that its proprietary AI algorithm can analyse 250 call-related parameters in just 2 milliseconds
read more
Airtel is taking a major step in addressing the issue of spam and scam calls with the introduction of an AI-powered feature for its subscribers. Users will now see a ‘Suspected Spam’ banner on their screens when receiving calls flagged by Airtel’s in-house AI system as potential spam. SMS messages containing harmful links will also be marked as suspicious to protect customers from falling prey to scams.
This new feature will be automatically enabled for both prepaid and postpaid users without any extra charges. While smartphones are ready for this upgrade, Airtel is still working on adapting the technology for feature phones.
Airtel’s move comes after years of unsuccessful attempts by telecom operators, tech companies, and regulators to tackle the problem of unwanted calls and messages. Previous efforts, like setting up a Do-Not-Disturb (DND) registry, have not been enough to curb the increasing nuisance.
Gopal Vittal, Bharti Airtel’s Managing Director and CEO, expressed that after spending the last year developing a comprehensive solution, the company is launching India’s first AI-powered spam-free network. He believes this will provide much-needed relief to customers from the constant barrage of intrusive communications.
AI detecting scam callers
Airtel’s approach to detecting spam relies on subscribers ignoring calls or SMS messages labelled as ‘Suspected Spam.’ The company hopes this will reduce the frequency of fraudulent communication as users become more aware.
The system works by passing every call and SMS through two filters — one at the network level and another at the IT systems layer. Airtel claims that its proprietary AI algorithm can analyse 250 call-related parameters in just 2 milliseconds. These indicators include the velocity and frequency of calls, device changes, unanswered calls, complaint patterns, and more.
The AI model is also designed to cross-check links in SMS messages against a central database of blacklisted URLs and monitor suspicious device activities like frequent IMEI changes. To address privacy concerns, Airtel has clarified that the AI model does not read SMS content but looks for specific markers indicating spam.
Currently, Airtel’s AI model has a 97 per cent accuracy rate for spam calls and a 99.5 per cent accuracy for spam SMS messages. Vittal stated that the system is continually improving, and the AI model is expected to close the small gap in accuracy over the next few weeks.
AI to the rescue
In response to the surge of spam communications, the Telecom Regulatory Authority of India (TRAI) had directed all telecom operators to deploy AI and machine learning-based detection systems. Despite these efforts, experts argue that AI alone may not be enough to combat the issue unless serious actions are taken against repeat offenders.
A recent proposal by TRAI suggested differential tariffs for users making over 50 calls or sending more than 50 SMSes a day, targeting heavy spammers. Airtel’s Vittal has also indicated that the company is open to sharing data with regulators to help identify and prevent scams before they start, as every scam begins with a call.
Airtel’s new AI-powered initiative marks a significant step in tackling the persistent problem of spam and scam calls, and while it may not be perfect, it promises to offer better protection to millions of subscribers.