← All articles
Guides

TRAI Fourth Amendment Regulations 2026: The AI Complaint Systems Checklist for BPOs

TRAI now lets AI agents run telecom complaint centres around the clock, as long as a human is reachable 16 hours a day. Here is what to check before you switch on a Human + AI model.

The short answer

On 2 November 2026, TRAI amended its telecom complaint-redressal rules to allow AI complaint handling systems and automated AI agents in complaint centres, provided a human is reachable at least sixteen hours daily under the TRAI AI customer support mandate. However, the system must run under “appropriate human oversight AI”; a phrase the regulation leaves undefined. Here is what to check before implementing a Human + AI BPO model.

What did the TRAI Fourth Amendment Regulations 2026 change for AI complaint handling systems?

Complaint centres move from a 16-hour operating window to round-the-clock availability, with AI complaint handling systems, automated software, or human representatives (or a hybrid Human + AI BPO model) now explicitly permitted to deliver support.

Here is how the rules compare before and after the Fourth Amendment:

BeforeAfter 2 November 2026
Operating window16 hoursRound the clock
Who may deliver supportHuman representativesHuman representatives, automated systems, AI agents, or a combination
Minimum human coverageImplicit (the 16-hour window)At least 16 hours daily, explicit
DisclosureNot specifiedCaller must be told when talking to an AI agent
Definition: Human + AI BPO

Human + AI BPO, the BPO where AI makes the calls and your people step in when it matters. More terms in our Human + AI BPO glossary.

What’s still required under the TRAI AI customer support mandate for human agents?

At least sixteen hours of live human assistance a day is mandatory, regardless of how advanced your AI customer support automation runs around it.

What does “appropriate human oversight AI” actually require in practice?

The regulation doesn’t explicitly define it. Until further regulatory clarification lands, the safest reading is an active Human + AI BPO operation: oversight that can be demonstrated on a live call with seamless AI-to-human call handoffs, not just documented in policy files.

What must callers be told regarding mandatory AI disclosure?

Callers must be clearly informed that they are speaking to an AI agent, a mandatory AI disclosure condition attached directly to the telecom regulatory compliance guidelines.

The AI customer support compliance checklist for contact centres this quarter

  • Can you show, on a real live call, what the human agent sees during an AI-to-human call handoff?
  • Does your 16-hour human assistance floor align with peak call volume, or just meet the TRAI AI customer support mandate minimum?
  • Is AI disclosure included across all call scripts and IVR prompts, not just sampled QA logs?
  • Who signs off on appropriate human oversight AI standards for your support desk, and is it documented for audit readiness?

Run the handover test on your own desk this week; ask to see the screen.

Key takeaways

  • AI agents, automated AI complaint handling systems, or a hybrid Human + AI BPO architecture are now explicitly allowed in telecom complaint centres under the TRAI Fourth Amendment Regulations 2026 (notified 2 November 2026).
  • Under the TRAI AI customer support mandate, a human agent must still be reachable for at least sixteen hours a day; the rest of the clock can run on AI customer support automation.
  • Callers must be explicitly told when they are interacting with an AI voice bot or automated agent.
  • “Appropriate human oversight AI” is a strict regulatory condition of permission, not merely a best-practice recommendation.
  • The only practical way to test oversight compliance without a legal team is to inspect the seamless AI-to-human call handoff interface when a caller is escalated.

Sources: TRAI, Telecom Consumers Complaint Redressal (Fourth Amendment) Regulations, 2026, notified 2 November 2026; Medianama, “TRAI mandates 24x7 complaint centres; human help for 16 hours”, 2 November 2026.

Frequently asked questions

Does the TRAI Fourth Amendment Regulations 2026 apply to every telecom company in India?

Yes; it is a binding TRAI regulation governing every licensed telecom operator’s customer complaint centre and AI complaint handling system.

Does “appropriate human oversight AI” mean a supervisor must monitor every automated call?

Not explicitly; the term is undefined in telecom regulatory compliance text, which makes establishing operational AI-to-human call handoffs and audit logs essential.

Can a telecom support desk rely entirely on AI customer support automation for part of the day?

Yes, provided the mandatory 16-hour live human assistance window is maintained daily, mandatory AI disclosure rules are followed, and seamless escalation paths are enabled.

Is the Human + AI BPO model mandated only for voice calls?

The regulation covers the complaint centre function broadly across channels; however, voice interactions present the most critical test for AI-to-human call handoffs and compliance.

O
Oriserve
AI for BFSI · OriserveOriserve · Reviewed by

Oriserve builds the outcome-execution platform for contact-centre processes, AI agents that run collections, renewals, retention and support calls, with a person on the exceptions.

ShareinX

Hear an AI agent handle a real call, in 30 seconds.

Get a call →or book a full demo