Is AI Cold Calling Legal? TCPA, DNC and AI Disclosure in 2026

Is AI cold calling legal? The short answer is yes, AI cold calling can be done legally in the United States, but only when it is built on the same rules that govern any outbound call, plus a few extra ones written specifically for automated and AI-generated voices. The longer answer is that the legal picture has several moving parts: the Telephone Consumer Protection Act (TCPA), the federal and state Do Not Call (DNC) registries, AI and robocall disclosure rules, consent requirements, and calling-hour windows. This article walks through each one in plain language so you can understand where the lines are.

This is general information, not legal advice. Rules change and your situation is specific, so confirm your program with qualified counsel before you dial. That said, ColdCalls.ai is designed from the ground up for TCPA-compliant AI calling, so the controls described below are built in rather than bolted on.

The TCPA is the foundation

The TCPA is the main federal statute behind almost every cold-calling rule you have heard of. It restricts the use of automated dialing systems and prerecorded or artificial voice messages, sets out consent standards, and creates a private right of action with statutory damages that typically run from $500 to $1,500 per violating call. Because damages are per call, a sloppy automated campaign can become very expensive very quickly. The practical takeaway is that scale without controls is the real risk, not AI itself.

Federal and state Do Not Call lists

The National Do Not Call Registry lets consumers opt out of most telemarketing calls. Many states also run their own DNC lists with separate rules. A compliant program scrubs every number against these lists before dialing and keeps an internal suppression list of anyone who has asked you to stop. Key points:

• Scrub against the federal registry and any applicable state registries on a current basis, not from a stale file.
• Honor company-specific do-not-call requests immediately and permanently.
• Maintain a documented process so you can show how and when scrubbing happened.
• Treat reassigned and ported numbers carefully, since a number that was once safe to call may not be today.

ColdCalls.ai performs real-time DNC scrubbing before a call is placed, so suppressed numbers never get dialed in the first place. Learn how that fits into a full workflow on our AI cold calling software page.

AI voice and robocall disclosure rules

In 2024 the FCC made clear that calls using AI-generated or cloned voices fall under the TCPA's restrictions on artificial or prerecorded voice calls. In plain terms, regulators treat a synthetic AI voice the way they treat any other artificial voice, which means the consent and disclosure expectations apply. The compliant response is simple and honest: disclose that the caller is an AI. A good program states clearly, early in the call, that the prospect is speaking with an automated AI assistant and gives them an easy way to opt out or reach a human.

This is where a disclosure-first design matters. ColdCalls.ai discloses that it is an AI on every single call, by default, with no option to hide it. That is a deliberate choice. Transparency is both the ethical position and the safer legal one.

State rules are tightening

Several states have added their own requirements on top of the federal baseline, and they are getting more specific about AI. Texas, for example, passed legislation (often referenced as SB 140) that addresses AI and automated calls, including expectations around prompt disclosure of an automated or AI caller near the start of the conversation. Other states regulate calling windows, registration of telemarketers, and the content of opening statements. Because requirements vary, a serious program configures behavior by jurisdiction rather than assuming one national standard.

• Some states require disclosure within the first several seconds of a call.
• Some require telemarketer registration or bonding.
• Some narrow the permitted calling hours beyond the federal window.

Consent and calling hours

Consent is the other pillar. Depending on the type of call and the relationship, you may need prior express consent or, for certain marketing using an artificial voice, prior express written consent (PEWC). Calling someone with an existing business relationship is different from cold outreach to a purchased list, and the safer path is to know which bucket each contact falls into. On timing, the federal baseline limits telemarketing calls to between 8 a.m. and 9 p.m. in the recipient's local time, and some states are stricter. Calling outside those windows is a common, avoidable violation.

ColdCalls.ai lets you configure calling hours per time zone and per jurisdiction, so the system simply will not dial outside an allowed window. Combined with consent tracking, that removes two of the most frequent sources of complaints.

So, can you do this legally?

Yes. The pattern that keeps AI cold calling on the right side of the line is consistent: scrub against DNC lists in real time, disclose the AI clearly and early, respect consent rules, stay inside permitted calling hours, and keep records that prove you did all of it. None of these are exotic. They are the same disciplines good human calling teams already follow, made automatic. A compliance-first AI SDR is, in many ways, easier to keep compliant than a room full of people, because the rules are enforced by software on every call instead of depending on memory and mood.

If you want to see how the controls are configured in practice, our pricing page lays out what each plan includes, and our guide to making AI cold calls turns these rules into a step-by-step workflow. Again, this is general information and not legal advice, so review your specific program with counsel before launch.