The Telecom Regulatory Authority of India (TRAI) has mandated that all access providers implement an Unsolicited Commercial Communication (UCC) detect system which uses artificial intelligence (AI) and machine learning (ML). It will be responsible for identifying and taking action against individuals or organisations sending commercial communication without proper registration as per the guidelines stated in the Telecom Commercial Communication Customer Preference Regulations, 2018 (TCCCPR-2018). Those who fail to register with access providers and utilise 10-digit mobile numbers to send commercial communications via messages or calls are referred to as unregistered telemarketers (UTMs).
TRAI has been implementing multiple measures to combat the issue of UCC, which has led to a decrease in complaints against registered telemarketers (RTMs). However, despite the measures taken by telecom service providers (TSPs), UTMs continue to engage in UCC activities. They often employ deceptive tactics like sending fraudulent messages with misleading links and phone numbers. Customers are tricked into giving out sensitive information, which can result in financial losses for the victims.
Access service providers have been urged to deploy detection systems based on what is suitable and feasible for them. However, as UTMs constantly develop new techniques to send unsolicited communications, the existing UCC detection systems implemented by access service providers cannot always catch these messages.
To ensure consistency in the implementation of UCC detect systems, TRAI has instructed all Access Providers to deploy an AI/ML-based system. The system should be able to continuously adapt and handle new signatures, patterns, and techniques employed by UTMs. Additionally, access providers have been directed to share intelligence regarding UCC through a distributed ledger technology (DLT) platform with other providers.
The system ensures enhanced accuracy as the AI/ML algorithms can analyse vast amounts of data and learn from patterns. This helps enable the UCC detect system to accurately identify and differentiate between legitimate and unsolicited communications. This reduces the chances of false positives and ensures that genuine messages are not mistakenly marked as spam.
Furthermore, access providers are required to ensure that the UCC detect system can identify bulk senders of UCC who are not adhering to the regulations. All access providers must comply with these directives and provide an updated report on the actions taken within 30 days.
The system also increases the efficiency of our resource utilisation. With the ability to automate the detection process, the AI and ML-based system can significantly reduce the manual effort required for monitoring and filtering unwanted communications. This frees up valuable human resources, allowing them to focus on more critical tasks.
Countries across the world are ramping up efforts to combat spam messages and calls. For instance, the Singapore Infocomm Media Development Authority (IMDA) announced that organisations that use alphanumeric Sender IDs to send SMS are now required to register with the Singapore SMS Sender ID Registry (SSIR). As OpenGov Asia reported, all non-registered SMS are labelled as “Likely-SCAM”. This functions similarly to a spam filter or spam bin. Consumers that get non-registered SMS labelled as “Likely-SCAM” are advised to exercise caution. If unsure, consumers are encouraged to check with family and friends. This will improve IMDA’s overall resilience against scams. As of January 2023, over 1,200 organisations have already registered with SSIR, using more than 2,600 SMS Sender IDs.
