Airtel taps AI to combat India’s rampant spam calls problem

Bharti Airtel, India’s second-largest telecom operator, said on Wednesday it’s rolling out an AI-powered spam detection solution to all its customers as it vowed to curb the rising spam calls and messages in the world’s most populous nation. 

The company, which counts Google among its backers, announced the launch of what it claims is India’s first network-based, AI-powered spam detection system. The feature will alert Airtel’s customers in real-time to suspected spam calls and text messages, automatically activating for all users at no additional cost. Airtel has about 387 million wireless subscribers in India. 

Indians have long been grappling with a relentless barrage of spam. A recent survey by startup LocalCircles found that a staggering 95% of users report receiving unwanted calls daily, with most facing at least three such intrusions. The government’s “Do Not Disturb” (DND) registry has proven largely ineffective, with 90% of registered users still plagued by spammers, the survey found. 

Gopal Vittal, Managing Director and CEO of Bharti Airtel, said the telecom operator spent 12 months building the new spam detection system. “Spam has become a menace for customers,” he said.

The new system employs a dual-layer protection mechanism, filtering communications at both the network and IT systems layers. Airtel’s data scientists developed a proprietary algorithm that analyzes various parameters such as caller usage patterns, call frequency, and duration in real-time, the company said. 

This AI-powered solution processes a staggering 1.5 billion messages and 2.5 billion calls daily – equivalent to 1 trillion records processed in real-time. Airtel’s new system has been able to identify 100 million potential spam calls and 3 million spam SMSes originating every day, the company said. 

Beyond call and message filtering, the AI system also scans SMS content for malicious links, cross-referencing against a centralized database of blacklisted URLs. This additional layer of protection aims to prevent users from accidentally clicking on suspicious links, a common vector for fraud and phishing attempts.

The system can also detect anomalies such as frequent IMEI changes, which are often indicators of fraudulent behavior, the company said.

India’s telephonic spam problem extends beyond mere annoyance. There’s also a marked rise in Indians reporting calls from scammers impersonating government authorities, delivery services, and credit providers.