A call made using an autodialer or a prerecorded or artificial voice; some are lawful, but the term usually refers to unwanted or illegal automated calls.
Technically, a robocall is any call that uses automated dialing equipment or a prerecorded or artificial voice. That includes legitimate uses such as pharmacy reminders, school closure notices, and fraud alerts from a bank, many of which are lawful when the caller has the required consent or an exemption applies.
The problem calls are those made without consent or for fraudulent purposes: impersonation scams, fake debt collection, bogus warranty offers, and similar schemes. These campaigns typically rely on high call volumes, short durations, spoofed or rotating numbers, and a chain of wholesale providers that makes the source hard to find.
Regulators address robocalls on several fronts. The TCPA governs consent and calling practices, the Truth in Caller ID Act addresses spoofing, and FCC rules under the TRACED Act require providers to authenticate calls, file robocall mitigation plans, respond to tracebacks, and block certain traffic. State attorneys general also enforce their own laws and coordinate closely with federal agencies, so a single campaign can draw attention from several regulators at once.
Why it matters to carriers
Every provider that carries voice traffic is expected to take reasonable steps to keep illegal robocalls off its network. Failing to do so can lead to removal from the Robocall Mitigation Database, which forces downstream carriers to stop accepting your traffic. Detecting robocall patterns early is far cheaper than responding to enforcement after the fact.