Someone Cloned Your Voice — What the Law Actually Says

Someone cloned your voice — what the law actually says. Yes, you can sue, but your strongest legal claims will come from state law, not federal intellectual property protections.

In Lehrman v. Lovo, Inc. (S.D.N.Y., July 10, 2025), a federal court dismissed most trademark and copyright claims brought by voice actors whose voices were allegedly cloned and marketed by an AI company, ruling that federal IP law offers little recourse for voice imitation alone. The court held that trademark law requires "secondary meaning" and that extending protection to voices alone would turn the Lanham Act into an impermissible "general right of persona." Copyright also does not protect voice imitation: under 17 U.S.C. § 114(b), a sound recording’s copyright does not cover independent fixations that merely simulate or mimic the original sound, and the human voice itself is not copyrightable because it is not fixed in a tangible medium.

While federal claims faltered, the court allowed state right-of-publicity claims under New York Civil Rights Law §§ 50 and 51 to proceed, finding that the statute’s "digital replica" provision could plausibly apply to AI voice clones even without visual likeness. This means individuals in New York — and in other states with similar publicity rights — may have a viable path to sue for unauthorized voice cloning, including claims for injunctive relief, damages, and unjust enrichment. The court further suggested that the ongoing operation of an AI model trained on a person’s voice could constitute a continuing violation each time it generates new audio, potentially tolling the one-year statute of limitations under NY law.

For businesses using AI voice technology, this underscores why consent is non-negotiable. At My AI Call Center, we treat AI-generated voices as artificial voices under the TCPA, requiring prior express consent for any outbound call — a practice aligned with both legal compliance and ethical outreach. We never use voices without verified permission, and we validate list source and consent records before launching any campaign, ensuring that every interaction respects both regulatory standards and individual rights. As state laws like Illinois’ BIPA and Tennessee’s ELVIS Act continue to evolve, the message is clear: the safest and most defensible use of voice AI begins with clear, documented consent.

The Legal Paths That Work: Publicity Rights, BIPA, and the ELVIS Act

If someone clones your voice without permission, your winning claims live in state law — not federal intellectual property law. The Lehrman v. Lovo decision in the Southern District of New York made that clear when it dismissed most federal trademark and copyright claims by two voice actors, while letting state-law claims move forward.

Right of publicity under New York law. The Lehrman court allowed claims under NY Civil Rights Law §§ 50 and 51 to proceed, holding that the statute's "digital replica" provision can plausibly cover AI voice clones even without a visual likeness. According to Skadden's analysis, the court also rejected the argument that the statute protects only deceased individuals, and it suggested that a model trained on a person's voice may commit a continuing violation each time it generates new audio — potentially extending the one-year statute of limitations.

Illinois BIPA: the most direct right to sue. Illinois's Biometric Information Privacy Act explicitly covers voiceprints, requires prior written notice and consent before collecting biometric data, and includes a private right of action — making it the most direct answer for Illinois residents, per legal analyses of voice cloning regulation. California's CCPA/CPRA adds consumer rights over voice data, including deletion and opt-out rights.

Tennessee's ELVIS Act. Enacted March 21, 2024, the ELVIS Act treats voice simulations as property, protecting "sound in a medium that is readily identifiable and attributable to a particular individual, regardless of whether the sound contains the actual voice or a simulation of the voice," as the Minnesota Barrister explains. It provides post-mortem protection for 10 years if the voice is commercially exploited by the state, otherwise 2 years after death.

Contract and consumer protection claims. In Lehrman, breach of contract and consumer protection claims survived dismissal. Notably, per Fredrikson & Byron's alert, one narrow copyright claim also survived: promotional use of an original recording beyond its license scope. The court held that electronically negotiated contracts — even platform messages using pseudonyms on Fiverr — can satisfy the Statute of Frauds.

Your practical checklist before suing:

• Check your state's specific protections — they vary significantly by jurisdiction, and some states rely on common law rather than statutes.
• Preserve written agreements and platform communications, which may support contract claims.
• Act promptly, though continuing-violation doctrines may extend filing deadlines in some states.

For businesses, the lesson is prevention: consent is the central legal issue, and publicly available audio does not grant permission to clone a voice. That is why My AI Call Center reviews list sources and consent records before any campaign launches, and treats AI-generated voices as artificial voices requiring prior express consent under the TCPA. If you run outbound calling campaigns, plan your campaign with compliance built in from the start — managed campaigns against approved, permissioned lists start at 9¢ per connected minute.

What to Do Right Now: Preserve Evidence and Act Promptly

If you believe your voice has been cloned without permission, taking immediate action can preserve your legal options. Start by gathering and securing any written agreements, emails, or platform messages related to how your voice was obtained or used — electronic communications like those on Fiverr can satisfy legal requirements for contracts under New York law, even when using pseudonyms according to recent court interpretations. These records are critical for breach of contract or consumer protection claims, which have proven more viable than federal intellectual property arguments in early AI voice cloning cases.

Next, investigate your state’s specific laws before proceeding, as protections vary widely. Illinois residents benefit from BIPA’s private right of action for biometric voice data, while Tennessee’s ELVIS Act explicitly treats voice simulations as property as defined in state legislation. New York allows right-of-publicity claims under Civil Rights Law §§ 50 and 51, which carry a one-year statute of limitations — though courts have suggested that ongoing AI generation may constitute a continuing violation, potentially extending that window based on judicial reasoning in the Lehrman case. Acting promptly remains essential, but understanding whether your state recognizes a continuing-violation doctrine could impact timing.

• Save all correspondence, including voice recording requests and usage limitations discussed
• Document where and how your voice was shared or licensed, if at all
• Note the date you first discovered the unauthorized AI voice clone
• Preserve screenshots or logs of the AI-generated content in use
• Avoid contacting the alleged infringer directly without legal counsel

Businesses like My AI Call Center build compliance into their operations by treating consent as non-negotiable — verifying list permissions and consent records before launching any campaign — a practice directly supported by legal guidance requiring prior express consent for AI-generated voices under regulations like the TCPA as noted in industry analyses. For individuals, this same principle applies: your voice is not public domain simply because audio of it exists online. Protecting your rights starts with evidence, awareness of local laws, and timely action.

The Business Side: Why Consent Records Decide Who Gets Sued

If someone clones a voice, the lawsuits usually follow the consent trail — and for businesses running AI voice campaigns, that trail is built long before the first call goes out. The Lehrman v. Lovo litigation shows why: the voice actors' strongest surviving claims were contract and state-law claims, while the company allegedly repurposed recordings obtained under the pretense of "academic research purposes only" and marketed clones under invented names like "Kyle Snow" (Skadden's analysis).

The lesson for organizations is simple: publicly available audio does not grant permission to clone a voice. Legal analysis of AI voice cloning regulation is blunt on this point — relying on implied or assumed consent creates exposure under privacy, biometric, and publicity laws (Resemble AI's legal overview). A voice recording sitting on a public platform is not a license.

For outbound calling, the regulatory picture is equally clear. The FCC has clarified that AI-generated voices in robocalls are treated as artificial voices under the TCPA, which means prior express consent is required before those calls are placed, with substantial penalties for violations (regulatory tracking on AI voice cloning). Courts are also reading statutes expansively: the Lehrman court reasoned that laws should be construed "to encompass future developments and technological advancements" (Skadden).

So what separates the companies that stay out of court from the ones that end up in a 60-page federal opinion? Their records. Businesses that verify list source and consent documentation before launching voice campaigns can answer the two questions that decide most disputes:

• Where did this list come from? Bought lists without clear permission records are a liability, not an asset.
• Can you produce the consent record for each contact — what they agreed to, and when?
• Does the campaign honor state-specific rules on calling windows, disclosures, and opt-outs?
• Can recipients reach a human or opt out by keyword, with requests logged and honored immediately?

The stakes are growing alongside the technology. The AI voice cloning market passed $2.1 billion in 2023, growing at over 28.4% annually (market analysis), and the FBI has formally warned that AI voice impersonation attacks are increasing. Regulators are paying attention.

This is why My AI Call Center checks list source and consent records before any campaign launches, and tells clients plainly when a list won't support the campaign. Managed outbound calling against approved, permissioned lists starts at 9¢ per connected minute — plan your campaign review and know your compliance posture before you spend anything.

Run Compliant AI Calling Campaigns Without the Legal Risk

Most lawsuits over AI voice cloning trace back to the same root cause: someone used a voice without permission. In Lehrman v. Lovo, the voice actors' contract and consumer protection claims survived precisely because the company allegedly obtained recordings under false pretenses, restricting use to "academic research purposes only" before marketing commercial clones under pseudonyms, according to Skadden's analysis of the decision. Structured, permissioned campaigns are built to avoid exactly that conduct.

The legal research is blunt about where the risk sits. "Publicly available audio does not grant permission to clone a voice," and relying on implied or assumed consent creates exposure under privacy, biometric, and publicity laws, as one legal overview of AI voice cloning regulation puts it. At the federal level, the FCC treats AI-generated voices in robocalls as artificial voices under the TCPA, requiring prior express consent and carrying substantial penalties for violations.

That reality is why My AI Call Center checks list source and consent records before any campaign launches. Bought lists without clear permission records are flagged, and in most cases declined — the same discipline the Lehrman court rewarded when it held that even informal platform messages can form enforceable contracts. The process mirrors what Fredrikson & Byron attorneys describe as the emerging compliance baseline: state publicity and consumer protection laws pose significant risks precisely when consent is missing.

Every compliant campaign shares the same guardrails:

• Consent and list review before launch — no list runs without verified permission records or documented review
• AI disclosure on every call — recipients can ask if the call is AI-assisted, request a human, or opt out
• Keyword opt-outs (STOP, REVOKE) honored immediately, with DNC requests carried across all campaigns
• Full outcome logs — disposition codes, per-call notes, and opt-out/DNC records routed back to the client

The stakes are rising alongside the technology. The AI voice cloning market passed $2.1 billion in 2023, growing at over 28.4% annually, and the FBI has formally warned that AI voice impersonation attacks are increasing. The 2024 robocalls using a cloned Joe Biden voice in New Hampshire prompted state investigations — a reminder that regulators act fast when synthetic voices meet unconsented audiences.

The lesson from the case law is simple: consent is the central legal issue, and it is knowable before launch or it is not. A campaign that confirms list source, discloses the AI on every call, honors opt-outs immediately, and keeps complete logs never engages the conduct that lands AI voice users in court. Nothing launches until you approve — script, disclosure, opt-out handling, and escalation path included. Campaign requirements still vary by location, industry, and consent status, so obtain appropriate legal guidance before launch.