Call centres
Recording and consent
Which parties must be told, why it depends on where the caller is, and what that means for how the system is configured. For another implementation reference, see this Monitask guide.
The question is where the other party isThe first question
Not where your call centre is. Jurisdictions differ on whether one party to a call knowing about a recording is enough, or whether every party must be told and must agree.
In the United States this varies by state, and the rule that generally matters for a given call is the one covering the person you are speaking to. A centre in one state calling customers across twenty is operating under twenty answers.
Several jurisdictions attach a private right of action, meaning the exposure is the person on the call rather than a regulator, and the amounts are set by statute rather than by harm.
Inbound and outbound are not the same
An inbound caller who hears an announcement and stays on the line has done something. An outbound call reaches somebody who did not choose to be in the conversation, and the announcement has to come before anything worth recording is said.
Which means the configuration differs by direction as well as by jurisdiction, and a single global setting is the arrangement most likely to be wrong somewhere.
What has to exist operationally
- An announcement configured per jurisdiction and per direction.
- A record of which announcement played on which call.
- A route for somebody who declines: a way to continue without recording, or a stated consequence.
- A retention period, applied automatically.
The third is the one usually missing. An announcement with no alternative is a notification rather than a consent mechanism, and in an all-party jurisdiction that distinction matters.
Callers hear the announcement configured for their location, and which announcement played is recorded against the call.
Agents see which of their calls were recorded and which were later reviewed.
The agent's side is a separate questionFrequently forgotten
Recording an employee at work engages employment and data protection rules that are distinct from the caller's consent, and in some jurisdictions requires notice, a stated purpose, or consultation with employee representatives before deployment.
Two different questions, two different sets of advisers, and answering one does not answer the other.
Agents are told what is recorded, why, who can hear it and how long it is kept, before any recording begins.
Nothing is recorded from an agent's machine beyond the call and its metadata.
Retention
A recording kept indefinitely is an obligation that grows: it is discoverable, it may be subject to access requests, and it has to be secured for as long as it exists.
Set a period tied to why the recordings are made, apply it automatically, and document the reasoning. Keeping everything because storage is cheap is a decision, and it should be made knowingly rather than by default.
Cross-border
A call between two countries may engage both. Where recordings are stored, and who can reach them, becomes a question in its own right.
None of this is exotic and all of it is answerable in advance. It is answerable considerably less comfortably after a complaint.
Not advice
This page describes the shape of a legal question, names no rule, and should not be relied on. The position differs between countries and between states and it changes. Take advice covering everywhere your callers are.
Where to start
Write down where your callers actually are, by volume. That list is the scope of the legal question, and it is usually shorter than people assume: most centres find eighty per cent of calls sit in a handful of jurisdictions.
Take advice on those first and configure them properly, rather than waiting for an answer covering everywhere.