Who decides a call is 'spam'? Inside the systems behind caller IDs

Your phone may warn you about spam before you answer. But who decides which calls get that tag, and on what basis? Read on

The Telecom Regulatory Authority of India (Trai) on Thursday asked the Ministry of Electronics and Information Technology (MeitY) to authorise it to take action against caller identification and call management applications such as Truecaller.

 

The development has brought attention to how these platforms identify spam calls and whether they should come under a formal regulatory framework.

 

MeitY Secretary S Krishnan said the ministry was examining Trai's request and was in discussions with the Department of Telecommunications (DoT). "They have reached out to us to seek powers to regulate call-management apps like Truecaller. We are consulting the DoT on granting powers, and we will have to see legally what to do about it," Krishnan said on the sidelines of the CII GCC Summit.

Why is this a matter of concern

The proposal comes after banks raised concerns that calls originating from 140- and 1600-series numbers were being marked as spam or blocked on caller ID platforms, which has affected communication with customers.

 

Officials said lenders also flagged instances where customers missed loan recovery calls or alerts related to suspicious transactions because these numbers had been classified as spam.

 

According to a report by Business Standard, Trai has not sought direct regulatory powers over these applications. Instead, it wants to be designated as an "authorised agency" under the Information Technology Act so that it can act against violations by such platforms.

 

Caller identification apps function as intermediaries under the IT Act and are not directly regulated under the Trai Act.

 

The proposal has also drawn criticism from sections of the telecom industry. Mahesh Uppal, director at Com First India, told Business Standard that extending Trai's powers beyond telecom services could set a wider precedent."

 

It's a bad idea for the simple reason that any extension of Trai powers to online applications is a thin end of the wedge, and therefore a problem. This may set a precedent for regulating other apps," he said.

How Truecaller, Airtel and Google identify spam calls

Truecaller, on its website, says it combines user feedback, automated detection systems and internal analysis to identify spam numbers. Users can manually report calls or SMSes as spam, and repeated reports contribute to a number's spam score.

 

The company says spam labels are not assigned only based on user reports. Its systems also examine calling patterns, the frequency of reports, abnormal calling behaviour and other signals before deciding whether a number should be classified as spam.

 

The platform, as per its website, claims to continuously reviews these signals, which allows spam labels to change if activity around a number changes.

 

For businesses using verified calling services, Truecaller says legitimate companies can request a review if they believe their numbers have been incorrectly marked. The platform also allows businesses to register official numbers so that customers can identify verified callers.

Airtel relies on network intelligence

Airtel's spam detection system works differently because it operates at the telecom network level.

 

The company, on its website, says its AI-powered solution analyses multiple parameters, including calling behaviour and network traffic patterns, to identify suspected spam calls and SMSes before they reach customers.

 

According to Airtel, the system processes billions of records every day and classifies suspicious communication in real time. Spam alerts are displayed to subscribers without requiring them to install a separate application.

 

The service also covers calls originating from over-the-top (OTT) communication platforms, where technically feasible.

 

The company has said its system does not access the content of calls or messages and uses metadata and network-level indicators to identify suspected spam.

Google's caller ID combines business data and user reports

Google's Phone app also offers caller identification and spam protection on supported Android devices.

 

According to Google, caller ID works by matching incoming numbers with publicly available business listings and other verified sources to display the identity of businesses when information is available.

 

Its spam protection feature combines information from Google's databases with reports submitted by users. Numbers reported by multiple users may be displayed with a spam warning during incoming calls. Users can also report or block numbers directly from the Phone app, helping improve spam detection over time.

 

Google has also introduced AI-powered scam detection for supported Android devices.

 

The feature, according to Google, analyses conversation patterns during calls processed on the device and alerts users if it detects characteristics commonly associated with financial scams.

 

Google says the processing happens on-device and audio is not stored or sent to its servers.

Jio focuses on customer awareness

Reliance Jio has primarily focused on consumer awareness and cybersecurity guidance to help users identify fraudulent calls and messages.

 

The company, according to its website, advises subscribers to avoid sharing banking credentials, verify unknown callers independently and report suspected fraud through official channels.

 

Unlike Airtel, Jio has not announced a network-wide AI spam detection system that automatically labels incoming calls for all subscribers.

 

The government's review of Trai's proposal could determine whether caller identification platforms remain governed only as intermediaries under the IT Act or come under an additional oversight mechanism. At the same time, the discussion has also brought attention to the different methods used by caller ID platforms, telecom operators and smartphone providers to identify suspected spam calls before they reach users.