Skip to main content
June 25, 2026 · By Piyush Sahoo · Updated November 3, 2026 Answering machine detection (AMD) is a call-analysis feature that listens to the first seconds of an answered outbound call and judges whether a person or a machine picked up. Your application then branches: connect someone, leave a message, or hang up. The verdict is a probability, so wrong answers are part of the design. This guide explains how detectors decide, why a headline accuracy figure tells you little, and how to measure the error that matters on your own traffic. It also compares the abandoned-call rules in India, the US, and the UK, because each one counts the same mistake differently.
Key takeaways
  • AMD starts where call signalling stops. The network tells you a call was answered. Only the audio can suggest who or what answered it.
  • Accuracy is the wrong first question. Two detectors can both score 92% while one drops ten times as many people. Ask for the false-machine rate instead.
  • Background detection doesn’t remove the wait. It starts your call flow at once, but an agent held for the verdict still stays quiet until the verdict arrives.
  • Phones now answer for people. iPhone, Pixel, and Truecaller screening can all put a spoken assistant in front of the person you called.
  • India’s denominator differs. TRAI’s 2018 criterion measures abandoned calls against attempted calls, while the US and UK measure against live answers.

What does answering machine detection decide?

AMD answers one narrow question after a call connects: is this a person? Ringing, busy, and no-answer are different outcomes, and they arrive as call progress and hangup signals. A voicemail system looks like an ordinary answer to the network, so the detector has to work from sound. Result labels vary by product. Typical results cover a person, a machine, a fax, or an unknown answer, and some say how a greeting ended. Three common uses follow from that verdict:
  • Protect agent time. Connect only answered humans to a live agent.
  • Protect AI spend. Keep a voice AI agent from talking to a recording.
  • Deliver a message. Wait for the greeting to end, then leave a voicemail.
How often a machine answers depends on your traffic, so measure it on your own calls.

How does a detector tell a person from a machine?

Most detectors weigh a few signals in the opening seconds. People tend to say something short and wait for a reply. Recordings tend to keep talking because nothing is listening. Illustrative timelines of five answers over eight seconds: a person's short hello, a person's long hello, a person who waits three seconds, a long voicemail greeting ending in a beep, and a phone call screener's spoken prompt. A duration rule reads the long hello and the screener as machines. Illustrative timelines of five answers over eight seconds: a person's short hello, a person's long hello, a person who waits three seconds, a long voicemail greeting ending in a beep, and a phone call screener's spoken prompt. A duration rule reads the long hello and the screener as machines. The last row matters for Indian traffic. A transcript-based check can only match phrases in languages its speech recognition handles. A timing rule ignores language, but it still assumes people everywhere greet at a similar length. Neither assumption should ship without a test on your own audience.

Why does a 92% accurate detector still fail?

Accuracy blends two different mistakes into one number, and those mistakes have very different costs.
  • False machine. A person is classified as a machine. You hang up on them or play them a recording.
  • False human. A machine is classified as a person. An agent or AI talks to a recording until someone notices.
The chart shows two invented detectors on the same traffic. Both are right on 920 of 1,000 answered calls. Bar chart of two illustrative detectors at 92 percent accuracy on 1,000 answered calls with 250 people and 750 machines. Detector A drops 5 people and passes on 75 machines. Detector B drops 50 people and passes on 30 machines. Bar chart of two illustrative detectors at 92 percent accuracy on 1,000 answered calls with 250 people and 750 machines. Detector A drops 5 people and passes on 75 machines. Detector B drops 50 people and passes on 30 machines. Detector A wastes more agent time. Detector B discards 50 of the 250 people you worked to reach. When machines dominate a list, a detector can also look strong simply by leaning toward “machine”. On this traffic, one that always says “machine” scores 75% and never reaches a person. Published figures don’t settle the question, because the conditions differ. Vendors quote numbers from different traffic, and most are self-reported, so don’t rank products on them.

Does background detection remove the wait?

Only part of it. Detection needs audio from the answered call, which often takes a few seconds. Very short timeouts tend to hurt accuracy. The design choice is what the person hears during that time. Three illustrative timelines with a verdict at about four seconds. Blocking detection leaves silence until the call flow starts. Background detection with an immediate agent greeting fills the audio sample with the agent's voice. Background detection with a held agent runs a wait or hold audio until the agent greets at the verdict. Three illustrative timelines with a verdict at about four seconds. Blocking detection leaves silence until the call flow starts. Background detection with an immediate agent greeting fills the audio sample with the agent's voice. Background detection with a held agent runs a wait or hold audio until the agent greets at the verdict.
  • Blocking detection suits flows with no live conversation, because the branch is simple and nobody is waiting to talk.
  • Background detection with an immediate greeting sounds fast, but it damages the sample. The person answers over the agent, and the classifier loses the “hello, then pause” shape it depends on.
  • Background detection with a held agent is usually the right design for voice AI. You still pay for the verdict in seconds, so size the timers deliberately and add a fallback.
A fallback covers the case where no verdict arrives in time. If the choice is between a person left on hold and an agent talking to voicemail, the person on hold is usually the worse outcome. Let the agent start.

What changed when phones began answering for people?

Some “machines” are now assistants with a person behind them. That breaks the old assumption that a machine answer means nobody is there. A screening prompt is long, continuous speech, so a duration rule tends to call it a machine. The person may join seconds later, which is why an automatic hang-up on that verdict can lose a real contact. We found no reliable public figure for how many Indian calls reach voicemail or a screener. Don’t import a US voicemail rate into an Indian campaign plan. Measure it on a labelled sample, as described below.

How do you measure AMD on your own calls?

You can’t find false machines by reading AMD results alone. Ofcom made this point in 2010: the equipment can’t identify its own false positive as a call picked up by a live person. So you need to hear what happened after the verdict.
  1. Keep the call open. Run a small test where a machine verdict doesn’t end the call, so the audio after the verdict still exists.
  2. Record lawfully. Follow your consent and retention duties. Compliant call recording pipelines covers the design.
  3. Label by listening. Mark each call as person, voicemail, screener, or network announcement, without looking at the verdict.
  4. Split the results. Report by country, operator, language, device group, and time of day where you can. Tuning on a handful of devices can mislead you once you deploy widely.
  5. Tune one timer at a time. Then repeat the test, because a change that saves people can let more machines through.
Plan for enough live answers. If you label 300 and find 6 false machines, the rate is 2%, but the 95% Wilson interval runs from about 0.9% to 4.3%. That sample can’t show you’re under a 3% limit. These are real calls to real people, and they cost money. Use your own Auth ID and Auth Token, confirm your list and consent, and start small.

Which AMD setup fits your campaign?

Start from what should change when a machine answers. Sometimes nothing should, and then AMD only adds a wait and a way to fail. Decision map from the question of what should happen if a machine answers. Nothing different: skip AMD and play the same short message. Don't spend a seat: use the fast verdict for live agents with a target for dropped people. Don't start the bot: hold the AI agent until the verdict. Leave a message: wait for the greeting to finish. Costly calls should never be auto-dropped on the verdict alone. Decision map from the question of what should happen if a machine answers. Nothing different: skip AMD and play the same short message. Don't spend a seat: use the fast verdict for live agents with a target for dropped people. Don't start the bot: hold the AI agent until the verdict. Leave a message: wait for the greeting to finish. Costly calls should never be auto-dropped on the verdict alone. Two checks apply to every branch. Decide who owns a late or missing verdict, and decide what your dialler does next. Ofcom’s current policy treats a repeat silent or abandoned call within 72 hours as an aggravating factor, unless a live agent was guaranteed. A false machine verdict looks like voicemail in your logs, so guaranteeing an agent on the retry is a sensible habit anywhere.

How do regulators treat AMD mistakes?

Regulators mostly don’t regulate the detector. They regulate what happens to the person who answered, and a false machine can count against you. Rules change often, so confirm the current text and take legal advice for your campaign.

The denominator changes your result

Take 1,000 attempts that reach 300 live people, with a detector that drops 5% of them. That is 15 people, and nothing else goes wrong.
  • Against live answers, as in the US rule and Ofcom’s formula, the rate is 5%. That is above the US limit of 3%.
  • Against attempted calls, as in TRAI’s 2018 criterion, the rate is 1.5%, which is below 3%.
Don’t treat the second number as a safe margin. It only shows that a figure from one country can’t be read across to another. The UK shows how this debate moved. In its 2008 statement, Ofcom worked through a scenario where a 3% false positive rate on machine verdicts, added to a 2% base of abandoned calls, produced a rate of 4.85%. Its current policy, in force since March 2017, names no specific rate and says the old 3% figure was never a safe harbour. A call with no message counts as a silent call, which sits outside the rate formula, and the policy names AMD as one cause. Ofcom’s January 2026 reminder says it can still act on these calls, with penalties of up to £2 million.

India: three layers of text

The 2018 regulations. TCCCPR 2018 defines abandoned calls as above and tells access providers to report entities above 3% abandoned, or above 1% silent, over 24 hours. Read the definition of a silent call closely. TRAI means a very short call designed to prompt a callback. Ofcom means a call where the person gets no message. Dead air from a slow detector fits Ofcom’s meaning, not TRAI’s. The February 2025 amendment. The Second Amendment replaced regulation 4 with a duty to notify your originating access provider in writing, in advance, about auto dialler or robo-call use and its purpose. We found no amendment deleting the 2018 definition or the 3% criterion, but we also found no consolidated text. Treat their current force as a question for your provider or counsel. The September 2026 amendment. The Third Amendment defines an A2P call as one started by an application, software system, or automated platform without direct human dialling. Four points from it affect AMD users:
  • Declaration. Senders must declare A2P use and CLI ranges in advance, and undeclared calls count as unsolicited.
  • Live agents don’t exempt you. TRAI considered excluding calls with a human agent and kept its definition, so a predictive dialler that connects agents appears to be in scope.
  • Transactional calls count. TRAI says the rules apply to all A2P calls, not only commercial ones.
  • A per-minute charge. TRAI’s press release describes a termination charge of up to ₹0.05 per minute between operators, with designated series exempt. Ask your provider how this reaches you. Minutes spent talking to a machine may matter more once it applies.
The notification starts the declaration and charge 60 days after Gazette publication. It is dated 18 September 2026, so they can’t start before mid-November 2026. We couldn’t confirm the Gazette date, so check it against your launch date. TRAI’s own data shows why this matters to anyone running automated calls. In April 2026, the amendment’s explanatory text says, 15,736 of 20,591 disconnections for these violations were SIP or PRI connections. One open question has no answer in the text we read. India’s definition says “live agent”, so a recorded message or an AI agent may not satisfy it. Raise this with your provider or counsel before you rely on either. Vobiz’s India calling regulations, UCC guide, and 140 and 160 series guide cover the wider requirements.

US: a two-second clock, under review

47 CFR 64.1200(a)(7) and the FTC’s 16 CFR 310.4(b)(1)(iv) both start the clock at the person’s completed greeting. A detector that waits for silence spends that time first. Bar chart of the US two-second limit after a completed greeting, with three illustrative settings. Waiting for 700 milliseconds of silence leaves 1,300 milliseconds. Waiting for 1,200 milliseconds leaves 800. Waiting for 2,000 milliseconds leaves nothing. Bar chart of the US two-second limit after a completed greeting, with three illustrative settings. Waiting for 700 milliseconds of silence leaves 1,300 milliseconds. Waiting for 1,200 milliseconds leaves 800. Waiting for 2,000 milliseconds leaves nothing. The chart shows illustrative detector settings, not measured verdict times. The remainder still has to cover your webhook (callback) round trip and the agent connection. The FCC has proposed deleting its abandonment limits in a notice published in December 2025. Kelley Drye’s summary covers the proposal. We found no final order as of 3 November 2026, and the FTC’s separate rule isn’t part of that FCC proposal.

How Vobiz handles AMD

Vobiz machine detection is a set of parameters on the Make Call API, not a separate endpoint. It applies to outbound calls placed through that API.
  • Background results. Set machine_detection_url and optionally machine_detection_method, and Vobiz sends webhooks (callbacks) with the result. The payload includes Machine, IfMachine, CallUUID, and an Event of MachineDetection.
  • Reporting. A call ended by machine_detection=hangup closes with hangup cause 9100, Machine Detected, which separates it from other outcomes in your data.
  • Greeting-end timing. In VobizXML, <Wait> with silence moves on once a greeting stops, and beep waits for a voicemail beep. The two are mutually exclusive, and the docs describe both as best-effort, so pair them with a fallback.
  • AI agents. Tuning for AI voice agents gives balanced, high-accuracy, and low-latency starting profiles. It also gives the gating steps: hold the agent, branch on Machine, and treat a late callback as a human.
  • Working example. The LiveKit AMD agent classifies the pickup before the conversation starts, and leaves a recorded message on voicemail.
  • Audio for agents. Audio streams and WebSocket streaming connect call audio to your agent. Vobiz also has Vapi and LiveKit integration guides.
  • Campaign sizing. Model channel needs in the CPS and concurrency calculator. The outbound campaign guide covers running at volume.
Calls placed with these settings are real and billed to your account. See the Vobiz pricing page for current rates, and confirm numbers and consent before you run a test.

Frequently asked questions

It listens to the first seconds of an answered outbound call and judges whether a person or a machine picked up. Your application then acts on that verdict. It might connect an agent, start an AI agent, leave a message, or end the call.
Published figures vary widely, but vendors measure them differently and most are self-reported. A single accuracy number also hides which mistakes occur. Measure how many live people your setup wrongly marks as machines, using your own destinations and call types.
It usually means a live person was classified as a machine. The call may be dropped or sent a recording, so this is normally the costliest error. The opposite error, often called a false negative, sends a machine to your agent or AI.
Yes. The detector needs audio before it can decide, which often takes a few seconds. Running detection in the background lets your call flow start immediately, but an agent held for the verdict still waits.
Sometimes, but a duration rule can read a spoken screening prompt as a voicemail greeting, even though a person may join moments later. Test screened calls separately, and treat a machine verdict on them as uncertain until you have measured it.
We found no TRAI rule that bans AMD itself. The rules govern the calls around it, including advance notice of auto dialler use and limits on abandoned calls. TRAI’s September 2026 amendment adds an A2P declaration. Confirm the current text and take legal advice for your campaign.

Further reading on Vobiz

Sources

Regulators and law Product documentation (read November 2026) News and legal commentary

Build outbound on Vobiz

Provision a number and test machine detection on your own outbound calls.