Rebuilding trust through intelligent network infrastructure

The telecommunications sector faces a credibility crisis. Industry-wide Net Promoter Scores rank among the lowest across all sectors, and in numerous markets, subscribers face relentless waves of fraudulent and spam communications. A 2024 survey by Age UK revealed a stark reality: nearly one-fifth of residents aged 50 and above—approximately 4.9 million individuals—avoid answering their phones due to concerns about scams and fraud.

Deploying AI-driven infrastructure offers carriers a pathway to address these core issues and recover lost ground. According to Philip Otley, Managing Partner for Telco, Infrastructure, Media & Entertainment at HTEC, "it can bring systematic intelligence directly into network infrastructure rather than [using] bolt-on solutions…[so]…CSPs can reclaim value in communications. This is fundamentally different from previous technology waves that simply increased speed or capacity."

Airtel's pioneering approach

Airtel has emerged as a global leader in this space, launching what it claims is the industry's first network-based spam detection system in late 2024. The platform automatically notifies users of potential threats by employing AI to examine network activity and identify suspicious calls, messages, and links instantaneously.

The World Economic Forum's February 2025 white paper on artificial intelligence in telecommunications highlighted Airtel's achievement, noting that the system processes "2.5 billion calls and 1.5 billion messages daily, it successfully identified close to 1 million spammers every day within the first two months." The dual-layer architecture combines network and IT infrastructure to deliver broad protection. By examining caller behavior patterns, the AI system identifies suspected spam without requiring customer intervention, delivering immediate safeguards while preserving privacy.

In May 2025, Airtel announced another milestone: the deployment of real-time, network-wide blocking of malicious websites across all communication channels—encompassing email, web browsers, over-the-top services including WhatsApp, Telegram, Facebook, and Instagram, as well as SMS.

Because fraud detection operates at the infrastructure layer of Airtel's mobile and broadband services, customers receive protection at no additional cost. This integration has substantially reduced the need for manual reporting and strengthened security. By June 2025, Airtel's AI-powered defenses had decreased customer financial losses by nearly 70 percent, with cybercrime incidents on the network declining by 14.3 percent overall. The Indian Cyber Crime Coordination Centre (I4C) under the Ministry of Home Affairs validated these results in September 2024.

The malicious domain blocking service initially rolled out across Haryana circle, covering most of the northern Indian state, with expansion to the entire country planned. Airtel continuously maintains a proprietary database of fraudulent domains, drawing intelligence from global sources including Mavenir, Openphish, Mindtest, Google, and Microsoft, with updates occurring every 24 hours.

The system's modular design permits straightforward integration with telecommunications infrastructure globally without requiring modifications. When a user follows a link, the system cross-references the domain against Airtel's database. Identified threats are immediately blocked. For domains not yet catalogued, the operator leverages partner APIs to evaluate risk factors based on user interaction signals and other indicators.

The research report examines how AI and autonomous networks are being deployed by operators including China Telecom, Comcast, Colt Technology Services, Deutsche Telekom, Fastweb, Google Cloud, Orange and Orange Business, Singtel, SK Telecom, Swisscom, Telefónica, Telenor, Verizon, Virgin Media O2, and Vodafone.

Source: Mobile Europe