The FCC just drew a line in the sand on artificial intelligence in telemarketing. In a Notice of Proposed Rulemaking (NPRM) released in March 2026, the Commission proposed the first federal rules specifically targeting AI-generated voice calls and text messages. If adopted, these rules would require a separate, distinct consent before any AI-generated communication reaches a consumer’s phone.
This is not a minor tweak to existing TCPA requirements. It is a structural change to how consent must be collected, stored, and proven for any company using AI in its outbound communication stack.
What the FCC Is Proposing
The NPRM lays out three core requirements for AI-generated calls and texts:
1. Mandatory AI Disclosure
Before obtaining consent, the caller or sender must clearly disclose that the consumer may receive calls or texts generated by artificial intelligence. This disclosure must be separate from the general TCPA consent language. It cannot be buried in a terms-of-service page or folded into a multi-purpose authorization.
2. Separate AI-Specific Consent
Consent to receive AI-generated calls or texts must be obtained independently from consent to receive human-initiated calls or texts. A single checkbox that covers “calls and texts from Company X” would no longer be sufficient if any of those communications are AI-generated. The consumer must affirmatively agree to receive AI communications as a distinct act.
3. Plain-Language Disclosures
The consent language must be written in plain, understandable terms. The FCC specifically called out the practice of embedding consent authorizations in dense legal paragraphs. Under the proposed rules, the AI disclosure must be “clear and conspicuous” and use language that a reasonable consumer would understand.
Why This Matters for Lead Generation
If you run a lead generation operation, or if you buy leads and contact consumers using any form of AI, this proposal should be at the top of your compliance agenda. Here is why.
AI Is Already Everywhere in Lead Gen
The days when AI in telemarketing meant a simple IVR system are over. Today, lead generation companies and their buyers are deploying:
- AI dialers that use predictive algorithms and voice synthesis to initiate and conduct calls
- AI-generated SMS campaigns where message content is created or personalized by language models
- Chat-to-call flows where an AI chatbot qualifies a lead and then initiates an outbound call
- AI voice agents that conduct entire sales or qualification conversations without a human on the line
Under the proposed rules, every one of these use cases would require separate, explicit consumer consent that specifically mentions AI. The general TCPA consent that most lead forms currently collect would not cover AI-generated communications.
The Consent Form Problem
Most lead forms today are built around a single consent disclosure. The consumer checks a box, the disclosure references calls and texts from one or more identified companies, and the lead is sold. That model has already been under pressure since the FCC’s one-to-one consent rule took effect in January 2025.
The AI consent proposal adds another layer. If you are generating leads for buyers who use AI dialers or AI texting, your consent forms now need to:
- Identify the specific company that will contact the consumer (one-to-one consent)
- Disclose that the consumer may receive AI-generated calls or texts
- Obtain a separate affirmative action for the AI-specific consent
- Present all of this in plain language that a reasonable consumer would understand
That is a significant change to form design, user experience, and data capture.
Dual Consent Creates New Proof Requirements
Here is the part that should keep compliance teams up at night. When consent is split into two distinct authorizations, you now need to prove both. In a TCPA dispute, it will not be enough to show that a consumer consented to receive calls. You will also need to show that the consumer separately consented to receive AI-generated calls.
That means your consent evidence must capture:
- The exact language of the AI disclosure as it appeared at the time of consent
- The consumer’s separate affirmative action (a distinct click, check, or signature)
- A timestamp for each consent action
- The technical context (IP address, device, browser session) tying the consumer to both consent events
If your current consent capture system records a single consent event, you will need to upgrade it to handle dual consent flows with independent evidence for each.
How This Connects to Existing TCPA Requirements
The AI consent proposal does not replace existing TCPA consent rules. It stacks on top of them. A company that wants to contact a consumer using an AI dialer would need:
- Prior express written consent under the existing TCPA framework (47 U.S.C. 227)
- One-to-one consent identifying the specific company (FCC one-to-one rule, effective Jan 2025)
- AI-specific consent under the proposed new rules
Each of these is a separate legal requirement with its own proof burden. Missing any one of them exposes the caller to statutory damages of $500 to $1,500 per call or text.
The math is straightforward. If you make 10,000 AI-generated calls without proper AI-specific consent, your exposure ranges from $5 million to $15 million. That is before class certification, which can multiply the numbers dramatically.
The Comment Period and Timeline
The FCC’s NPRM opens a public comment period during which industry participants, consumer groups, and other stakeholders can submit feedback. Based on the FCC’s recent rulemaking pace, final rules could be adopted within 6 to 12 months of the comment period closing, with an implementation window of 12 to 18 months after adoption.
That timeline means companies should not wait for final rules to start planning. The direction of travel is clear: the FCC views AI-generated communications as categorically different from human-initiated ones and intends to regulate them accordingly.
What About State Laws?
Several states have already moved ahead of the FCC on AI disclosure. Colorado’s AI Act requires disclosure of AI use in certain consumer interactions. California’s AB 2655 targets AI-generated content in political communications. Illinois’s AI Video Interview Act mandates disclosure when AI is used in hiring.
While none of these state laws directly mirrors the FCC’s proposed TCPA framework, they signal a broader regulatory trend. Companies that build AI consent capture into their lead flows now will be better positioned as state-level AI telemarketing laws inevitably follow.
How Consent Verification Platforms Need to Evolve
The AI consent proposal puts new demands on consent verification technology. Platforms like eConsent that capture and store consent evidence will need to support:
- Dual consent capture: recording two distinct consent events within a single lead session, each with its own disclosure language and consumer action
- Disclosure versioning: tracking which version of the AI disclosure was presented to each consumer at the exact moment of consent
- Granular evidence packaging: producing consent certificates that separately document general TCPA consent and AI-specific consent, with independent evidence chains for each
- Form audit capabilities: allowing compliance teams to verify that live lead forms meet both the one-to-one consent requirement and the new AI disclosure requirement
This is not a future problem. Companies that wait for final rules to begin building these capabilities will find themselves scrambling during the implementation window.
What You Should Do Now
1. Audit your AI usage. Map every point in your lead generation and contact workflow where AI is involved. This includes AI dialers, AI-generated SMS content, chatbot-to-call flows, AI voice agents, and any AI-based lead qualification or routing that results in outbound contact.
2. Redesign your consent flows. Begin planning consent form updates that separate AI-specific consent from general TCPA consent. Test user experience designs that present dual consent clearly without creating excessive friction.
3. Upgrade your consent evidence. Make sure your consent capture system can record and store two distinct consent events per lead, each with its own disclosure snapshot, timestamp, and consumer action evidence.
4. Review your lead buyer agreements. If you sell leads, your contracts with buyers need to address AI consent. Buyers who use AI to contact leads will need proof that AI-specific consent was obtained. If your forms do not capture it, your leads lose value and your indemnity risk increases.
5. Submit comments. If you operate in lead generation, file comments during the FCC’s public comment period. Industry input shapes final rules. Silence does not.
The FCC’s AI consent proposal is the most significant expansion of TCPA consent requirements since the one-to-one rule. Companies that treat it as a distant concern will find themselves in the same position as those who waited too long to implement one-to-one consent: behind, exposed, and scrambling. The ones that move now will be ahead of the market when the rules take effect.