Voice Cloning
How to Spot Scams & the Companies Fighting Back
By Mary Keane, CyberWyoming Cybersecurity Business Counselor with assistance from AI
Imagine your phone rings late at night. You hear your daughter’s voice—panicked. “Mom, I’m in trouble. Please, send money now.” Your heart races. It sounds exactly like her—same tone, same inflection, even her pet nickname for you. But it’s not her.
Welcome to the new world of AI voice cloning, where technology can mimic anyone’s voice from just a few seconds of audio. It’s impressive, and terrifying, all at once.
What Is Voice Cloning?
Voice cloning uses artificial intelligence to analyze and reproduce a person’s speech patterns, tone, and rhythm. With just a few samples—sometimes less than 10 seconds—AI can generate a digital version of your voice that sounds startlingly real. Like all powerful tools, its impact depends on who’s using it — and what safeguards are in place.
When It Helps: The Good Side of Voice Cloning
Voice cloning can be used for good, from restoring speech to people with medical conditions to saving time in film production. Here are two inspiring examples:
Giving Voices Back: Sarah Ezekiel was diagnosed with ALS and gradually lost her ability to speak. For years she relied on a robotic-sounding voice device, but in 2025, a company used just eight seconds of old home video to recreate her natural voice. When she began using the new voice, her students and friends said they recognized her again. Sarah described it as “getting a piece of myself back.” This story highlights the compassionate side of AI technology.
For more on how the ALS community is using voice banking, see https://www.youralsguide.com/voice-banking.html.
Helping Creators and Businesses: In Hollywood, voice cloning is used transparently and with consent. James Earl Jones provided consent for his voice to be used for Darth Vader beyond his lifetime so that Disney could continue to use his iconic voice in new Star Wars films. Another Star Wars actor, Mark Hamill, also licensed his voice for future use. After Val Kilmer suffered throat cancer, his voice was recreated with AI to complete his last film, Top Gun: Maverick.
Unfortunately, criminals are also exploiting this technology for scams, impersonation, and fraud. Here are some examples of how it’s being misused:
- The “Grandparent Scam 2.0”: A grandmother got a late-night call from her grandson’s voice claiming he was in jail and needed $9,000. The voice was most likely cloned from social media.
- The Fake CFO: A finance worker wired $25 million after a convincing AI video call from their ‘CFO.’ The voice and face were both synthetic.
- The Voice-Over Artists’ Nightmare: Two voice-over artists provided their voices for a free-lance project, and then they found their voices were being used for podcasts for an AI company.
How Ethical Companies Are Fighting Back
Fortunately, many legitimate voice-cloning companies are taking strong steps to prevent misuse. Here’s what some leading innovators are doing:
- Descript (Overdub): Requires users to record a verification script and undergo human review before cloning. Its policy explicitly bans cloning anyone else’s voice without permission.
- ElevenLabs: Implements consent verification, account checks, and partnerships with detection firms like Reality Defender to identify fake voices. It also bans political and explicit content.
- Resemble AI: Focuses on enterprise clients with consent-first creation and its own detection tool, Resemble Detect, to flag AI-generated audio.
- Respeecher: Serves film studios and advertisers under strict contracts, requiring explicit written consent and refusing unethical requests.
- VocaliD: Originally designed to help people with ALS and throat injuries bank their voices, emphasizing consent, accessibility, and privacy.
Common Safeguards These Companies Use
Most responsible AI voice companies share a set of core principles and controls:
- Consent verification: Users must record a verification phrase or sign agreements.
- Human review: Suspicious cloning requests are manually reviewed.
- Watermarking and detection tools: Audio is tagged or traceable to detect synthetic speech.
- Privacy compliance: Adherence to GDPR and CCPA ensures users can delete their data.
- Transparent policies: Public commitments against non-consensual or harmful uses.
Protecting Yourself and Your Business
Even as companies improve safeguards, human awareness remains your best defense:
- Pause before you act: Hang up and call the person back using a verified number.
- Create a family password: Choose a simple phrase everyone knows. If someone can’t say it, it’s likely a scam.
- Be cautious with what you post: Voice clips, videos, and podcasts can be cloned.
- Train employees: Include voice-cloning scams in cybersecurity awareness programs.
- Use multi-channel verification: Never approve financial actions based solely on a voice or video call.
WHAT TO LOOK FOR IN A RESPONSIBLE VOICE-CLONING VENDOR
Feature |
Consent verification |
Clear privacy policy |
Human review |
Watermarking/detection |
GDPR/CCPA compliance |
Why It Matters |
Prevents cloning without permission. |
Ensures voice data can be deleted. |
Detects suspicious or fraudulent requests. |
Helps track and identify fake audio. |
Ensures responsible data protection. |
The Takeaway
Voice cloning can give people their voices back—or steal them. The difference lies in consent, transparency, and ethics. While technology companies are building stronger safeguards, personal awareness is still the most powerful protection. Until every AI-generated clip is clearly marked, your best defense is to pause, verify, and remember your family password.
At CyberWyoming, we help small businesses and communities stay informed about evolving scams like this one.
