Consumer groups want probe of AI voice-cloning app Speechify over fraud concerns

The Consumer Federation of America is urging federal and state regulators to investigate whether an artificial intelligence company has failed to build adequate safeguards into one of the most powerful—and potentially dangerous—consumer AI technologies now widely available.

In a complaint filed Monday with the Federal Trade Commission and attorneys general across the country, the CFA and students from the UCLA School of Law's Information Policy Lab argue that AI voice-cloning platform Speechify enables convincing voice impersonations while doing too little to verify that users actually have permission to clone someone else's voice.

The complaint asks regulators to investigate whether the company's practices violate Section 5 of the FTC Act, which prohibits unfair or deceptive business practices, as well as similar state consumer-protection laws and digital forgery statutes.

Why consumer advocates are concerned

Voice cloning has rapidly evolved from a novelty into a practical tool capable of reproducing a person's speech from only a few seconds of recorded audio.

That technology has legitimate uses. People who have lost the ability to speak because of illness or injury can preserve or recreate their voices. Businesses use synthetic voices in customer service, audiobooks and accessibility tools.

But the same technology has become a favorite tool of scammers.

Fraudsters have used cloned voices to impersonate:

• Children supposedly calling parents after an accident
• Grandchildren claiming they've been arrested
• Bank employees
• Government officials
• Corporate executives authorizing wire transfers

CFA argues that companies offering these services should not simply rely on users checking a box stating they have permission to clone someone's voice.

"Speechify has made it remarkably easy for anyone to clone a person's voice while doing almost nothing to verify consent or prevent abuse," Ben Winters, CFA's director of AI and data privacy, said in announcing the complaint.

Billions lost to impersonation scams

The consumer group points to rapidly rising fraud losses.

According to FTC data cited in the complaint:

• Consumers reported losing $3.5 billion to impersonation scams in 2025.
• That figure is nearly three times higher than in 2020.
• Older Americans alone lost $445 million to impersonation scams during 2024.

While not all of those scams involved AI-generated voices, regulators increasingly warn that generative AI is making impersonation fraud far more convincing and much harder to detect.

Regulators have been sounding the alarm

The complaint builds on several years of growing concern inside the FTC.

In 2023, the agency launched its Voice Cloning Challenge, seeking technologies capable of authenticating voices, detecting AI-generated speech and limiting misuse before consumers become victims. FTC officials warned at the time that voice cloning could dramatically increase fraud against consumers while also threatening performers, artists and other creators whose voices could be copied without permission.

Consumer organizations have also urged the FTC to adopt rules that would allow regulators to pursue AI developers and service providers when they knowingly—or they should reasonably know—their technologies are facilitating fraud.

The broader debate: Who is responsible?

The Speechify complaint highlights a rapidly emerging legal question.

Traditionally, regulators have focused enforcement on the scammers themselves.

Consumer advocates increasingly argue that AI companies should also bear responsibility when they release products that make fraud easy while failing to implement reasonable safeguards.

Possible safeguards include:

• Strong identity verification before voice cloning
• Multi-factor consent from the person whose voice is being copied
• Digital watermarking of synthetic audio
• Detection systems that flag suspicious activity
• Limits on anonymous or mass-generated voice cloning

Technology companies generally respond that these tools have many beneficial uses and that responsibility lies primarily with criminals who misuse them.

Exactly where regulators draw that line could shape the next generation of AI regulation.

What consumers can do

Experts recommend taking several steps to reduce the risk of AI voice scams:

• Create a family "safe word." Agree on a phrase known only to close family members that can verify an emergency call is genuine.
• Slow down. Scammers rely on panic and urgency.
• Verify independently. Hang up and call the person back using a trusted phone number.
• Never send money immediately based solely on a phone call.
• Be cautious about publicly posted recordings. Podcasts, social media videos and other online recordings can provide source material for voice cloning.

What this means for consumers

The complaint is unlikely to produce immediate enforcement action, but it signals that consumer groups are shifting their focus beyond individual scammers to the companies that build AI tools.

If regulators agree, AI developers may be required to implement much stronger identity verification, consent procedures and fraud-prevention systems before consumers can use voice-cloning services.

That could make AI voice tools slightly less convenient—but potentially much safer.

Bottom line

AI voice cloning has moved from science fiction to consumer reality in just a few years. The same technology that can restore speech to patients or produce natural-sounding audio books can also make scams dramatically more believable.

CFA's complaint against Speechify reflects a broader shift in consumer protection policy: rather than pursuing only the criminals who exploit AI, regulators are increasingly being asked to examine whether the companies building these technologies are doing enough to prevent foreseeable harm.

If the FTC takes up the case, it could become one of the first major tests of how much responsibility AI platforms bear when their products become tools of consumer fraud.