Taking Card Payments Over the Phone: PCI DSS and the Voice System
A customer reads out sixteen digits, an expiry date, and a three-digit code. The agent types them into a payment page and the call continues. Nothing about that exchange feels like a security incident, and yet for a few seconds the business has handled cardholder data over a channel that was designed for conversation, not for payment.
Why the Phone Is the Weakest Link in Card Security
Every other payment channel has been engineered to keep card data away from people. Terminals encrypt at the point of entry, online forms post directly to a payment processor, and chip transactions never expose the full number. The phone is different: the number passes through a human ear, a microphone, a codec, a recorder, and often a spreadsheet maintained by someone in finance. Each of those is a place where data can be retained without anyone deciding to retain it.
The practical consequence is scope. If card numbers are spoken on a recorded line, the recording system may fall inside the scope of a payment card assessment, which turns a small compliance exercise into a large one. The work of voice payment compliance is therefore mostly the work of narrowing scope until the phone system is provably outside it.
| Approach | How it works | Residual risk |
|---|---|---|
| Agent takes the number and types it | Simple, no integration work | Number is in the recording, the agent's memory, and often a note field |
| Pause and resume recording | Agent pauses, takes details, resumes | Depends entirely on agent discipline and on the pause being audited |
| DTMF masking at the platform | Keypad tones replaced with silence or a flat tone | Customers who read numbers aloud instead of keying them are not covered |
| Payment IVR with no agent | Caller keys card details into an automated flow | Needs a clean handoff back to the agent and a separate recording policy |
| Payment link by SMS or email | Call centre sends a link, customer pays elsewhere | Slower; the customer may not complete it during the call |
The Controls That Actually Help
Mask the Keypad, Not Just the Words
Dual-tone multi-frequency masking intercepts the tones produced when a caller keys a card number and replaces them in the stored recording with silence or a flat tone. It is the only control that removes the data at the point of entry rather than relying on a person to behave correctly. The important detail is the instruction given to callers: if the IVR asks them to key the number but they read it aloud anyway, masking does nothing at all. Prompt wording has to be explicit and the agent script has to match it.
Make Pause and Resume an Audited Event
Where masking is not available, pause and resume is the fallback, and it fails quietly. Agents forget to resume, so the rest of the call disappears from the record; or they forget to pause, so the number lands in storage. Both are invisible unless someone checks. The control that makes it real is a report: every pause event with its duration, reviewed monthly, with calls that contain a payment reference but no pause flagged for listening.
Keep the Number Out of the Notes
The recording is the obvious risk. The ticket is the one people forget. Free-text fields accumulate card numbers, expiry dates, and the last four digits in combinations that are themselves identifying. A field-level rule that blocks long digit strings, plus a periodic scan of notes, closes a gap that no amount of recorder configuration will fix.
What to Write Down
- Which call flows can carry card data, and which are deliberately excluded
- Whether masking, pause and resume, or an automated IVR is used at each entry point
- How long recordings are kept, and what happens to a recording that contains card data
- Who can retrieve a recording, and what the retrieval log looks like
- How the agent script tells callers what to key rather than what to say
- The review that proves the control is working rather than merely configured
The Questions Auditors Ask
Assessments of voice payments tend to converge on a small set of questions. Is card data stored anywhere in the call path? Can you prove it is not? Who has access to the recordings, and is that access logged? What happens when an agent takes a payment without following the process? A business that can answer those four questions with evidence rather than assurance has done most of the work, and usually discovers along the way that the simplest fix is to stop taking card numbers on the phone at all.
Redstone Approach
Redstone IP-PBX deployments support pause and resume recording, per-extension recording permissions, and retrieval logging so that access to a recording is a recorded event rather than an assumption. Where card payments remain on the voice channel, deployments are scoped so the recorder is configured deliberately rather than by default, and the resulting configuration is documented for assessment.
Redstone Systems, Inc. founded in Delaware, USA in December 2002, has been the ODM vendor for many well-known communications companies, serving the Southeast Asian market. In 2020, Redstone Systems will begin to return to the North America market with its self-developed brand.
Redstone has a complete product line of intelligent voice gateways, providing IP-PBXs, analog VoIP gateways (FXS/FXO), digital VoIP gateways (E1/T1), border appliances, and session boundary controllers (SBCs).
With advanced technology in digital signal processor (DSP), speech coding and speech processing, as well as efficient operational tools such as cloud remote management, auto provisioning, Redstone gateways are widely used in markets of enterprise communications, cloud communications, call centers, operators’ IMS/SIP trunks, bringing users friendly, efficient and reliable communication experience.
For more information about Redstone, You can follow us on Facebook, Linkedin, and Youtube to be the first to get the latest news.

Current Location :

