Skip to content
Voice AI

Voicemail Detection for AI Calls: How Answering Machine Detection Works and What to Do When It's Wrong

How an outbound AI voice agent tells a person from a voicemail greeting: carrier-level and agent-level answering machine detection, what LiveKit and Twilio publish about it, why no vendor publishes a reliable accuracy figure, what to do on each outcome, what it costs, and the rules for leaving a message.

Jaffar Jawed · Co-Founder
Published
7 min read
  • Outbound Calls
  • Voicemail Detection
  • Telephony
  • Voice AI
In short

Quick answer

An AI voice agent detects voicemail with answering machine detection: it classifies the first seconds of an answered call as a person, a voicemail greeting, a phone menu or an unreachable line. No vendor publishes a dependable accuracy figure, so decide what happens on each result and treat uncertain ones as a person. Botcadence's outbound voice agent can hang up on voicemail.

How does an AI voice agent know it has reached voicemail?

Through answering machine detection, usually shortened to AMD. When an outbound call is answered, the system listens to the first seconds and classifies what it hears: a person, a voicemail greeting, a phone menu or an unreachable line. It can do this at the carrier, by timing the greeting, or inside the agent, by understanding the words. Neither is perfect. Twilio's documentation says its detection "will not always return the right answer", and LiveKit publishes no accuracy figure for its own. The practical job is to choose what happens on each result and to make a wrong guess harmless. Botcadence's outbound voice agent has optional voicemail detection on campaigns that hangs up when it hears an answering machine.

Key takeaways

  • A voicemail system answers the call just as a person does, so the phone network alone cannot tell them apart.
  • Two approaches: carrier detection based on audio timing, and agent detection that classifies the first thing said.
  • No vendor we read publishes a dependable accuracy figure. Expect errors both ways.
  • Decide five behaviors: person, voicemail, phone menu, unreachable, uncertain.
  • Carrier detection has a per-call price: $0.0075 on Twilio and $0.002 to $0.0065 on Telnyx.
  • A message left by an AI voice is still an artificial voice call under the TCPA, with the same consent and identification rules.

Why voicemail is hard to detect

LiveKit's outbound calling documentation notes that a voicemail system answers with the same signal as a person picking up. From the network's point of view the call connected. Everything after that is inference from audio:

  • A person usually says a short "hello?" and waits.
  • A voicemail greeting is longer, continuous, and ends with a beep.
  • A phone menu speaks in a different pattern again and waits for a choice.

Short custom greetings ("Hi, it's Sam"), people who answer with a long sentence, and carriers' call-screening features all blur the line.

Two ways to detect it

Carrier-level detectionAgent-level detection
How it decidesTiming and energy of the first audio: length of speech, pauses, a beepA model classifies the first utterance by what was said
Where it runsIn the carrier's network, before or as the call is bridgedIn the voice agent, on the first thing it hears
OutcomesTypically human, machine, fax, unknownLiveKit's are human, machine-ivr, machine-vm, machine-unavailable and uncertain
CostPer call: $0.0075 on Twilio; $0.002 standard or $0.0065 premium on TelnyxUsually part of the agent's normal processing
WeaknessShort greetings and long hellos confuse it; it can add a pause before the agent speaksIt needs to hear some speech first, and depends on the model

LiveKit's documentation says its detection runs once, on the first utterance. That is the approach Botcadence uses on outbound campaigns: the classifier listens to the opening words, with examples of common United States and India voicemail greetings to guide it.

What nobody publishes: accuracy

We looked for accuracy figures on the vendors' own documentation and found none we could rely on. Twilio states only that the result will not always be right. LiveKit lists the outcomes without a rate. Any percentage you see on a sales page should be read with the question "measured on which calls, in which country, on which carrier?"

Two kinds of error matter, and they cost different things:

ErrorWhat happensCost
A person is classified as a machineThe agent hangs up on, or leaves a message for, a live personA lost conversation and an annoyed contact
A machine is classified as a personThe agent talks to a greeting, then to silence after the beepWasted minutes and a garbled voicemail

Tune toward the error you can tolerate. For calls to your own customers, hanging up on a person is worse; lean toward treating uncertain results as human.

What to do on each result

ResultSensible behavior
PersonStart the conversation. Identify the business at once
VoicemailHang up and retry later, or leave one short, approved message
Phone menuHang up, unless the call is meant to reach an extension and the agent can navigate it
Unreachable or disconnectedMark the contact and stop retrying
UncertainTreat as a person: greet, and listen for a reply

On Botcadence, with voicemail detection switched on for a campaign the agent hangs up when it hears voicemail or an unreachable line. With it off, the agent is instructed to leave one short message. Detection is off by default and is set per campaign. Retries follow the campaign's rules: the first call plus up to three retries by default, one a day per contact, at least four hours apart.

Should the agent leave a message?

Arguments for hanging up: no risk of a clumsy half-message, no minutes spent on a greeting, and the contact sees a missed call from a number they can call back.

Arguments for a message: the contact learns who called and why, which raises callbacks for expected calls such as appointment confirmations.

If you leave one, keep it under twenty seconds: who is calling, why, and the number to call back. And remember the law. The FCC treats AI-generated voices as artificial voices under the TCPA, and its rules require an artificial voice message to state the business's identity at the beginning and give a phone number. Consent requirements apply to the call whether or not anyone answers. See are AI phone calls legal?.

What it costs

ItemList price, read November 12, 2026
Twilio answering machine detection$0.0075 a call
Telnyx answering machine detection$0.002 a call standard; $0.0065 premium
Retell AI batch calling$0.005 a dial
Time on voicemailBilled as call time on most platforms; on Botcadence at least 15 seconds for an answered call

A campaign of 2,000 dials with Twilio's detection adds $15. The larger cost is minutes spent listening to greetings when detection is off or wrong.

How to test it

  1. Call ten phones you control: some answered, some sent to voicemail, one with a short custom greeting, one with a carrier's default greeting, one with call screening on.
  2. Note the result for each and how long the agent took to speak.
  3. Answer with a long sentence ("Hello, this is Maria speaking, who's calling please?") and see whether you are treated as a machine.
  4. Check what was left on each voicemail.
  5. Repeat on each carrier you call most.

Where Botcadence fits

Botcadence's outbound campaigns include optional voicemail detection based on LiveKit's answering machine detection, with hang-up on voicemail, plus calling hours by time zone, limited retries and automatic opt-out when a contact asks not to be called. It does not publish an accuracy figure, and you should test it on your own contact list as described above. For keeping outbound numbers in good standing see how to reduce Spam Likely labels on AI calls, and for what the calls can do see the outbound sales voice agent page.

Questions readers ask

Voicemail Detection for AI Calls FAQ

Answering machine detection, or AMD, classifies an answered outbound call as a person, a voicemail greeting, a phone menu or an unreachable line, so the system can speak, hang up, leave a message or retry. It works from the first seconds of audio.

No vendor we read publishes a dependable figure. Twilio's documentation says its detection will not always return the right answer, and LiveKit lists outcomes without an accuracy rate. Test it on phones you control, on the carriers you call most.

For expected calls such as appointment confirmations, a short message can help: who is calling, why, and a number to call back. For other campaigns, hanging up and retrying is cleaner. Either way the consent and identification rules for artificial voice calls apply.

Carrier detection is charged per call: Twilio lists $0.0075 and Telnyx $0.002 standard or $0.0065 premium, on prices read November 12, 2026. Detection done inside the voice agent is usually part of its normal processing.

If a person is taken for a machine, the agent hangs up on them or starts a message. If a machine is taken for a person, the agent talks over the greeting and leaves a confused voicemail. Treating uncertain results as a person avoids the more costly error.

Sources

  1. LiveKit Docs: Answering machine detection (checked November 12, 2026)
  2. LiveKit Docs: Making outbound calls (checked November 12, 2026)
  3. Twilio Docs: Answering machine detection (checked November 12, 2026)
  4. Twilio voice pricing, United States (checked November 12, 2026)
  5. Telnyx Voice API pricing (checked November 12, 2026)
  6. Retell AI pricing (checked November 12, 2026)
  7. FCC: Declaratory Ruling FCC 24-17 (February 8, 2024) (checked November 12, 2026)
  8. 47 CFR 64.1200 (eCFR) (checked November 12, 2026)

Written by

· Co-Founder, Botcadence

Jaffar Jawed co-founded Botcadence in 2025 and leads product and go-to-market. He writes about AI voice agents, inbound answering workflows, lead qualification and the telephony stack behind real-time voice, including the company's open-source work on LiveKit noise and echo cancellation.

Get Started Today

See These Strategies in Action

Book a free demo and we will build a working AI agent for the channel this article covers, tailored to your business goals. No commitment required.

  • Voice AI plan $299 a month
  • Managed voice live in about 5 business days
  • Works with your existing tools

Ask AI about this page