Why It's Getting Harder to Tell — and Why the Law Is Stepping In

Not long ago, you could spot a robocall in the first two seconds — the flat tone, the awkward pauses, the scripted cadence. Today, AI voices respond naturally, handle interruptions, and sound convincingly human, which means your ear alone is no longer a reliable detector.

To make it harder, disclosure is inconsistent. Some AI systems clearly identify themselves as automated, while others stay silent or dodge the question entirely, leaving callers unsure who — or what — they're dealing with (Retell AI's analysis of AI phone call ethics). That gap is exactly what regulators are now closing.

In February 2024, the FCC unanimously ruled that AI-generated voices count as "artificial" voices under the Telephone Consumer Protection Act — meaning calls using them require prior express consent, no matter how human they sound (the FCC's ruling). Notably, there's no exception for conversational AI: even an agent that responds dynamically is treated the same as a prerecorded robocall (Thoughtly's breakdown of AI disclosure requirements).

A proposed FCC rule would go further, requiring callers to state at the start of every AI-generated call that they're using "AI-generated technology" — before any substantive conversation begins (Potomac Law's 2024 regulatory review). A final rule could arrive in 2026, and states aren't waiting:

• Utah gives callers a legal right to ask whether they're talking to AI and get a truthful answer.
• California prohibits bots from misleading people about their artificial identity.
• Colorado will require disclosure starting in 2026 unless the bot's nature is obvious to a reasonable person.

And enforcement shows this isn't theoretical. Lingo Telecom was fined $1 million in August 2024 for transmitting AI-generated robocalls that imitated President Biden's voice (reporting on the Lingo Telecom penalty), while the FCC proposed a $6 million fine in the related New Hampshire robocall case (details on the proposed penalty). Under the TCPA itself, violations run $500 to $1,500 per call with no cap — a 10,000-call campaign can mean $5 to $15 million in exposure (Henson Legal's AI voice compliance guide).

The direction is clear: disclosure is shifting from courtesy to legal obligation. That's why honest AI is built into how CallMyLeads answers calls — callers always know they're speaking with AI, and every caller can reach a human, switch to text, or book online. As the rules tighten, businesses that disclose upfront won't just stay compliant; they'll earn the trust that evasive systems lose.

Your Legal Right to Know: What Federal and State Laws Say

The question "Am I talking to a human or AI?" is increasingly answered by law, not guesswork. In February 2024, the FCC unanimously ruled that AI-generated voices are "artificial" voices under the TCPA, making AI robocalls without prior express consent illegal. A proposed FCC rule would go further, requiring callers to disclose at the start of every AI-generated call that they are using "AI-generated technology" before any substantive conversation begins.

For now, no final federal disclosure mandate exists — your rights depend on your state. A patchwork of state laws is filling the gap:

• Utah (effective May 2024): Regulated professions must proactively disclose generative AI; other businesses must disclose truthfully when asked — giving callers a legal right to ask and get an honest answer.
• California (Bot Disclosure Law, 2019): Prohibits using bots to mislead about their artificial identity to incentivize a transaction or influence a vote; disclosure serves as a legal defense.
• Colorado (effective 2026): Requires disclosure unless it would be "obvious to a reasonable person" they're interacting with a bot.
• Five more states — Hawaii, Idaho, Illinois, Massachusetts, and New York — have pending bills mandating chatbot disclosure or accountability for deceptive AI communications.

Enforcement is already active. The FCC fined Lingo Telecom $1 million for transmitting AI-generated robocalls imitating President Biden's voice, and proposed a $6 million fine in the New Hampshire case. TCPA violations carry $500–$1,500 per call with no cap — a 10,000-call campaign can mean $5–$15 million in statutory exposure. California's bot law allows up to $2,500 per violation, and Colorado's AI Act sets civil penalties up to $20,000 per violation.

For businesses, proactive disclosure is becoming both a legal shield and a trust-builder. CallMyLeads bakes this into every interaction — callers always know they're talking to AI, and every caller can reach a human, use text, or book online. Clear disclosure is a feature, not something hidden behind.

How to Check Mid-Call: Practical Ways to Verify You're Talking to AI

You don't have to guess whether the voice on the other end belongs to a person. A few simple moves, made mid-call, will tell you almost everything you need to know.

The most reliable test is also the simplest: just ask. Say "Am I speaking with an AI?" and pay attention to the answer. In Utah, the AI Policy Act gives callers a legal right to ask and receive a truthful answer, with regulated professions required to disclose proactively, according to legal analysis of state AI disclosure rules. Violations there carry administrative fines up to $2,500 each, per Thoughtly's breakdown of disclosure requirements.

Well-built systems don't wait for you to ask twice. Thoughtly's agents, for example, "clearly identify as AI when a caller directly asks whether they are speaking with a person or a machine," a behavior the company says is built into the platform by default. A system that dodges, deflects, or imitates a human when directly questioned is telling you something important about the company behind it.

Next, listen for the upfront disclosure. Honest companies increasingly lead with it, using scripts like: "You're speaking with an AI-powered assistant. If you'd like to speak with a person at any time, just let me know." Proposed FCC rules would require AI identification "before any substantive conversation begins," according to Henson Legal's compliance analysis. If you hear that disclosure in the first ten seconds, you're likely dealing with a business that takes compliance seriously.

Then test the human handoff. Ask to speak with a person, or request a text or online booking option instead. Legal experts note that offering a switch to a human agent boosts trust and reduces legal risk, and a handoff path is considered a practical compliance requirement. A trustworthy AI treats this as a normal request, not a dead end.

Here's a quick mid-call checklist:

• Ask directly: "Am I speaking with an AI?" — and expect a straight answer.
• Note whether the call opened with a clear AI disclosure.
• Request a human, a text follow-up, or an online booking link and see what happens.
• Watch for evasive answers or fake-human claims — both are red flags.

The stakes behind these checks are real. The FTC's September 2024 "Operation AI Comply" sweep established that failing to disclose AI-driven interactions can constitute an unfair or deceptive practice, per reporting on the enforcement sweep. And customers who discover mid-conversation they've been talking to undisclosed AI often feel deceived, leading to lost loyalty and reputational damage, as Retell AI's ethics analysis notes.

This is exactly why CallMyLeads builds disclosure into every call: callers always know they're talking to AI, and every caller can reach a human, switch to text, or book online. When a system handles your verification test gracefully, that's not a flaw in the AI — it's proof the business behind it has nothing to hide.

What Honest AI Looks Like — and What to Do If a Call Feels Deceptive

The best AI phone calls don't hide what they are — they announce it in the first breath. A gold-standard opening sounds like this: "Hi, this is [Company Name]. You're speaking with an AI-powered assistant. I can help you with [purpose]. If you'd like to speak with a person at any time, just let me know." That script pattern, recommended in compliance guidance, does three things at once: it identifies the AI, states its purpose, and guarantees a human path.

Proactive disclosure is rapidly becoming both a legal shield and a trust-builder. Proposed FCC rules would require AI identification "before any substantive conversation begins," with a final rule possibly landing in 2026. States aren't waiting — Utah fines non-disclosing businesses up to $2,500 per violation, and Colorado's AI Act carries penalties up to $20,000 per violation when it takes effect. Legal analysis also notes that offering a switch to a human agent boosts customer trust while reducing legal risk.

If a call feels deceptive, callers have real recourse:

• TCPA claims — statutory damages of $500–$1,500 per call, with no cap on total exposure.
• FTC complaints — the agency's "Operation AI Comply" sweep treats undisclosed AI interactions as deceptive practices.
• State attorney general complaints — AGs in Massachusetts, California, and Oregon have signaled active enforcement.

Enforcement isn't theoretical. The FCC fined Lingo Telecom $1 million for transmitting AI robocalls imitating President Biden's voice, and proposed a $6 million penalty in the same New Hampshire case.

This is why CallMyLeads treats disclosure as a feature, not a loophole. Its Honest AI approach means callers hear upfront that they're speaking with an AI assistant, can reach a human at any point, and can opt out immediately — the same model disclosure pattern regulators are moving toward. Every call also answers truthfully when a caller asks, "Am I talking to a machine?"

For businesses weighing AI answering, the lesson is simple: transparency now costs nothing and buys everything — compliance, caller trust, and a head start on rules that are coming regardless. The companies that disclose first won't be scrambling when the rules arrive; they'll already be trusted.