On 13th June 2023, TRAI issued a direction requiring all Access Providers to deploy Artificial Intelligence and Machine Learning (“AI/ML”) based Unsolicited Commercial Communication (“UCC”) Detect System to detect, identify and act against senders of Commercial Communication who are not registered in accordance with the provisions of Telecom Commercial Communication Customer Preference Regulations, 2018 (“TCCCPR”).
- As a part of its Direction, TRAI noted the following:
- There has been an increase in UCC calls and SMS from Unregistered Telemarketers (“UTMs”) using ten-digit mobile numbers.
- Use of fraudulent links and misuse of telephone numbers for trapping customers into sharing of their critical information often leading to financial loss to the customers.
- The inability of the current UCC detect systems to detect senders of such unsolicited communications.
- Constant evolution of SMS phishing (smishing) patterns, CTAs (call to action) and other frequently changing techniques by UTMs which due to their low activity span, are difficult to contain.
- In light of such evolving cases of misuse, the TRAI has recognised a need to use Al/ML techniques to deploy and constantly evolve the UCC Detect system capable of dealing with new signatures, new patterns and new techniques used by UTMs. Accordingly, there is a need to establish a UCC Detect System to detect, identify, and act against senders of Commercial Communication who are not registered in accordance with the provisions of the TCCCPR.
- TRAI has directed all Access Providers to:
- Deploy AI/ML based UCC Detect system capable of evolving constantly to deal with new signatures, new patterns and new techniques used.
- Ensure that UCC Detect System detect bulk UCC that is non-compliant with the TCCCPR.
- Ensure that UCC Detect System can –
- Undertake reputation-based analysis of the message sender, and is helpful in avoiding false positives, taking into account various factors like age of subscription, authentication at the time of subscription, address verification method, and SMS sending/ calling pattern;
- Share intelligence with other Access Providers using DLT platform.
- Ensure that action is taken by the Originating Access Provider (“OAP”) as per provisions of the regulations.
- Share information with concerned Law Enforcement Agencies, Ministry of Home Affairs, and Department of Telecommunications.
The introduction of AI/ML measures to combat UCC is a welcome step for the effective implementation of the TCCCPR. Over the past few years, the TRAI has been actively engaged in combating UCC, specifically through the strict implementation of registration requirements for telemarketers. The growing menace of bulk UCC, despite strict regulations for content and principal entity registration, is quickly escalating on a global scale. The Direction is extremely pertinent in light of the recent steps taken by the Singapore Infocomm Media Development Authority (“IMDA”). The IMDA requires entities sending SMS with alphanumeric Sender IDs to be registerd with the Singapore SMS Sender ID Registry (“SSIR”). Further, unregistered entities sending such SMS shall be labelled as “Likely-SCAM” in an attempt to preserve cyber health and protect consumers from online scams. Considering that previous attempts by TRAI utilising DLT/blockchain have not been able to keep up with the rise in UCC, incorporating AI/ML solutions to detect such communications could be extremely beneficial for consumers at large.
Press Release: https://pib.gov.in/PressReleaseIframePage.aspx?PRID=1932088
Link to IMDA circular: https://www.imda.gov.sg/how-we-can-help/anti-scam-measures#:~:text=This%20registration%20is%20to%20better,that%20may%20be%20online%20scams.&text=From%2031%20January%202023%2C%20non,as%20%E2%80%9CLikely%2DSCAM%E2%80%9D.