Using AI to answer incoming customer calls is not subject to TCPA consent requirements. The FCC's February 2024 declaratory ruling applied existing TCPA restrictions to outbound AI-generated voice calls — calls a business initiates to a consumer. That ruling does not govern calls consumers place to your business (Wiley Law, 2024). Here's what that distinction means in practice.
What the FCC's February 2024 ruling actually covers
The FCC voted unanimously in February 2024 to confirm that AI-generated voices used in outbound calls are subject to the same TCPA restrictions that already applied to artificial and prerecorded voice calls. If a business uses AI to dial customers — promotional messages, appointment reminders, marketing campaigns — those calls must comply with TCPA consent and opt-out requirements (Wiley Law, 2024).
This was significant for businesses running outbound calling operations. It closed a gap some had hoped existed — a belief that AI-generated voices might escape TCPA restrictions because they're dynamic rather than prerecorded. They don't. An AI voice on an outbound call is treated the same as a recorded message under the statute.
But the ruling targeted calls a business initiates to a consumer's phone. That's where the TCPA has always applied — and where it continues to apply with AI voices in the mix.
The distinction that matters: inbound versus outbound
The TCPA restricts calls made to consumers. It was designed to stop unwanted marketing calls reaching people who didn't ask for them. When a customer dials your number to book an appointment, ask about your service area, or request a quote, you didn't place that call. The current TCPA framework focuses on outbound calls — the inbound call answering context sits outside those consent requirements (Thoughtly, 2024).
An AI phone system answering your shop's incoming calls is not running an outbound campaign. It's answering a call the customer chose to make to your number. That's a fundamentally different legal category — and it's why the compliance picture for AI phone answering is much simpler than the headlines suggest.
Most news coverage about AI call violations focuses on outbound cases: robocall operations, AI-generated fundraising calls, lead-gen campaigns that contacted people who never opted in. Those violations carry real exposure — penalties per unauthorized outbound call with no statutory cap on total liability. But those cases have nothing to do with a plumbing shop in Kalispell that uses AI to answer after-hours calls from customers who dialed in.
| Scenario | TCPA Applies? | What's Required |
|---|---|---|
| AI makes outbound marketing calls to a customer list | Yes | Prior express written consent + opt-out mechanism + disclosure |
| AI sends automated texts to prospects or customers | Yes (TCPA text provisions) | Prior express consent; written consent for marketing texts |
| AI answers incoming calls from customers | Generally no | No TCPA consent requirement — inbound calls are customer-initiated |
| AI calls back a customer who left a voicemail | Generally no (return call context) | Best practice: identify the AI at call start |
What state laws add to the picture
Federal TCPA requirements don't cover inbound call answering — but some state laws add disclosure requirements that apply regardless of call direction. California's bot disclosure law requires businesses to disclose when AI is used to incentivize a commercial transaction in a real-time conversation. Utah's AI Policy Act requires disclosure when the caller directly asks whether they're talking to an automated system (Thoughtly, 2024).
State AI legislation is an active area in 2025 and 2026. For Montana-based service businesses primarily serving Montana customers, no state-specific AI phone disclosure mandate applies as of this writing. For businesses with customers in California — a Flathead Valley contractor fielding calls from out-of-state vacation property owners, for example — California's requirements are worth understanding.
Beyond state law, the FTC's authority over unfair and deceptive practices applies regardless of call direction. Using AI in a way designed to deceive callers into thinking they're speaking with a human — denying it's AI when a caller asks directly — could constitute an unfair practice under Section 5 of the FTC Act. That's a separate track from TCPA compliance, and it applies everywhere.
Best practice in 2026: disclose regardless of what's required
The most effective AI phone systems aren't the ones hiding what they are. Across systems deployed for Montana and Northwest service businesses, the approach that works is simple: the AI introduces itself clearly at the start of every call. 'Hi, this is [Business Name]'s scheduling assistant — how can I help you today?' Callers know within seconds they're speaking with an automated system.
Callers who understand they're talking to AI tend to be more efficient. They give the system exactly what it needs to route or book them. They don't wait on small talk. Transparent disclosure also eliminates any FTC deception concern — an AI that identifies itself isn't deceiving anyone, regardless of which state the caller is calling from.
What actually creates risk for service businesses
For a service business using AI to answer incoming customer calls, TCPA liability is not the exposure to worry about. The real risks are operational: an AI that misqualifies leads, drops escalations that needed a human, or creates a confusing experience for a caller trying to reach someone about an emergency. A frustrated caller who couldn't get through is more likely to leave a negative Google review than to file a federal complaint.
That changes if you also run outbound AI campaigns — text follow-ups to past customers, AI-initiated reminder calls to new leads, marketing calls to your contact list. If the same AI vendor powering your inbound answering also supports outbound functionality, know which mode you're running and what consent framework covers it. Different tools, different rules.