TRAI Enforces AI Spam Detection Rules: Telecom Operators Must Act on Flagged Calls

TRAI Mandates AI-Based Spam Detection Framework

The Telecom Regulatory Authority of India has introduced stringent regulations requiring telecom operators to deploy AI spam detection systems and take direct enforcement action against suspected spammers2

2

3

3

. The TCCCPR amendment marks a shift from passive spam warnings to active investigation and enforcement, even without consumer complaints. Telecom operators must now use AI and machine learning systems to identify suspected spammers based on calling patterns and behavioral analysis1

1

. During April-June 2026, operators received 721.79 billion incoming calls, with AI systems flagging 22.99 billion as suspected spam calls1

1

. The systems also flagged 1.44 billion of 836.31 billion incoming spam messages as unsolicited commercial communication.

How AI-Based Caller ID Systems Identify Spam

Telecom operators deploy sophisticated UCC_Detect systems that analyze multiple behavioral parameters to classify numbers as "Suspected high probability UCC CLI"3

3

. Airtel uses more than 250 derived signals to assess spam behavior, examining call frequency, call duration, usage patterns, suspicious web links, and device IMEI numbers1

1

. When AI spam detection systems identify a suspected spammer, operators display warnings such as "Airtel Warning: SPAM" or "Suspected SPAM" before users answer1

1

. Vodafone Idea's Voice Spam Detection system combines AI models with web scanning tools and user feedback to identify suspicious callers in real time1

1

. Pattern analysis happens through Distributed Ledger Technology platforms where operators share flagged numbers within two hours of identification3

3

.

Investigation and Enforcement Against Suspected Spammers

Under TRAI regulations, when five or more numbers linked to a sender are flagged within 10 days, operators must investigate and initiate graduated enforcement2

2

3

3

. For first-time violations, originating access providers must re-verify KYC verification within three business days3

3

. Second instances trigger physical KYC within five business days, and if misuse is confirmed, all telecom operators must bar outgoing services on the sender's telecom resources for 15 days3

3

. Repeat offenders face disconnection of all telecom resources including PRI/SIP trunks and SIMs for one year, regardless of whether those resources were used for spam calls2

2

. Operators must place violators under blacklisting spammers category for one year, preventing them from obtaining new telecom resources2

2

.

Consumer-Facing AI Tools Enhance Protection

Truecaller has introduced AI Call Scanner that analyzes a caller's voice during calls to determine whether it is human or AI-generated, currently available to beta users in India1

1

. The company's AI-powered Assistant can answer calls and ask callers their purpose, helping users decide whether to pick up1

1

. Google's Phone app offers Scam Detection on supported Pixel devices using Gemini Nano, an on-device AI model that flags potential scams during calls1

1

. Google processes conversations locally without storing or sending audio to servers, though the company acknowledges scam detection is not 100% accurate1

1

. However, TRAI regulations bar call management applications from tagging or blocking promotional telemarketing calls from designated series like 140xx, 1600xx, and 1601xx2

2

.

Lower Complaint Thresholds with AI Corroboration

The TCCCPR amendment reduces the complaint threshold from 5 to 3 unique recipient complaints within ten days when AI systems corroborate the suspected spam activity3

3

. In April 2026, AI systems flagged 771 crore calls and 48 crore SMS as suspected spam, while operators received 2,66,105 complaints with only 17,401 meeting the earlier threshold3

3

. This gap demonstrates why TRAI combined complaint data with AI outputs for enforcement. When the new threshold is met, originating access providers must immediately suspend outgoing services and start investigations3

3

. Subscribers wanting to block promotional calls can text BLOCK PROMO to 1909, which blocks promotional commercial communications while allowing transactional and service messages2

2

. The DLT platform facilitates information sharing between operators to track suspected spammers across networks. Terminating access providers can charge up to Rs 0.05 per minute for A2P calls under the amended rules2

2

.