CYBERSECURITY & ARTIFICIAL INTELLIGENCE

Artificial intelligence can now reproduce a person's voice convincingly enough to deceive friends, relatives and colleagues. Criminals are exploiting this technology to create fake emergencies, impersonate trusted people and pressure victims into sending money or revealing sensitive information. Understanding how these scams work—and how to verify someone's identity—can help protect you.

Imagine receiving a phone call from your mother, brother, close friend or business partner. The voice sounds familiar, the person knows your name, and the situation appears urgent. They claim they have been arrested, involved in an accident or stranded somewhere and urgently need money.

You recognise the voice, so you believe the story. But there is one problem: the person speaking may not be your relative or friend at all. A criminal could be using artificial intelligence to imitate their voice.

AI voice cloning is changing the way impersonation scams can be carried out. Criminals no longer have to rely entirely on pretending to be someone through text messages or using an unfamiliar voice. With suitable audio samples and voice-generation technology, they can create speech that resembles a real person's voice.

The US Federal Trade Commission (FTC) has warned that scammers can use voice cloning in family emergency schemes. The FBI has also documented malicious campaigns involving AI-generated voice messages used to impersonate trusted individuals and establish credibility with targets.

What Is AI Voice Cloning?

AI voice cloning is a technology that uses artificial intelligence to generate speech resembling a particular person's voice. Depending on the system and the available audio, a generated voice may imitate characteristics such as tone, accent, rhythm and pronunciation.

The technology has legitimate uses. It can help people with speech disabilities communicate, support voice production in entertainment and assist with authorised narration. The risk arises when someone uses it to impersonate another person without permission.

Criminals can potentially obtain voice samples from publicly available videos, social media posts, interviews, voice notes or other recordings. They can then use voice-generation tools to create audio that sounds like the person they intend to impersonate.

This does not mean every short recording is enough to produce a perfect copy. Results depend on the quality of the sample, the technology being used and other factors. However, the technology has advanced enough that people should not treat a familiar-sounding voice as conclusive proof of identity.

How Criminals Use Cloned Voices

1. Fake Family Emergencies

One of the most concerning uses is impersonating a family member who supposedly needs urgent help.

A scammer may call a parent and claim to be their child. The caller might say there has been an accident, they are in police custody or they urgently need money for medical treatment. A cloned voice can make the story feel more believable, especially when the victim is frightened.

The criminal may insist that the victim act immediately and avoid contacting anyone else. The aim is to prevent the victim from checking whether the emergency is real.

The FTC advises people to verify these claims independently instead of trusting the voice alone.

2. Impersonating Friends and Romantic Partners

Criminals can also exploit friendships and online relationships. Someone might receive a voice message that appears to come from a close friend asking for a loan, claiming to be stranded or requesting help with an unexpected problem.

In other cases, a criminal may combine a fake social media account, stolen photographs, convincing messages and a cloned voice to create a more believable identity.

The danger is not limited to people who are unfamiliar with technology. Anyone who trusts the person they believe is contacting them could be vulnerable to emotional pressure.

3. Business and Workplace Fraud

Voice cloning can also be used against businesses. A criminal may impersonate a manager, executive, supplier or business partner and request an urgent payment or confidential information.

For example, an employee could receive a call that appears to come from a senior manager instructing them to transfer money to a new account. If the employee believes the voice is genuine, they may follow the instruction without using the company's normal verification process.

This type of deception can overlap with business email compromise and other impersonation scams. Companies should never rely on a voice call alone to authorise unusual payments or changes to financial details.

4. Fake Government or Official Messages

Another risk involves criminals impersonating public officials or other influential people. An apparently familiar voice may be used to establish trust before the target is asked to disclose personal information, provide an authentication code or transfer money.

In a December 2025 public service announcement, the FBI described ongoing malicious messaging campaigns involving impersonation of senior US officials through text and AI-generated voice messages. The agency warned that criminals used these approaches to build trust and persuade targets to take actions that could expose sensitive information or assets.

The broader lesson is that an official-sounding voice—or one that resembles someone in authority—is not proof that a request is legitimate.

Why These Scams Can Be So Convincing

Traditional impersonation scams often give victims reasons to become suspicious: an unfamiliar voice, unusual wording or a message from an unknown account. AI voice cloning can remove one of those warning signs by making the caller sound like someone the victim already knows.

However, the technology is only part of the problem. Criminals also exploit human emotions and familiar social habits.

  • Fear: The caller claims a loved one is in danger.
  • Urgency: The victim is told there is no time to verify the story.
  • Trust: The voice resembles someone the victim knows personally.
  • Secrecy: The caller says not to tell other relatives, colleagues or friends.
  • Financial pressure: The victim is asked to send money immediately through a difficult-to-reverse payment method.

These tactics can make people act before thinking carefully. The most effective defence is to separate the emotional claim from the decision to send money or disclose information.

Can You Identify a Cloned Voice?

Sometimes an artificial voice may contain unusual pauses, unnatural pronunciation, inconsistent emotion or strange audio quality. But these signs are not reliable enough to establish whether a voice is genuine.

Modern generated audio can sound convincing, while genuine calls can also suffer from poor connections, background noise or unusual speech patterns. A person should not assume a call is safe simply because it sounds natural—or conclude that it is fake merely because it sounds unusual.

The FBI has warned that AI-generated content can be difficult to identify. For this reason, independent verification is more dependable than trying to detect every possible technical flaw in a recording.

How to Protect Yourself From AI Voice Scams

1. Hang Up and Call Back

If someone calls claiming to be a relative or friend who urgently needs money, end the call and contact that person using a phone number you already know is theirs.

Do not rely on a number supplied by the caller or automatically trust a familiar name displayed on your phone. Caller identification can be misleading, and a scammer may try to move the conversation to another number or messaging platform.

If the person does not answer, contact another trusted relative or someone who can independently confirm the situation.

2. Create a Family Verification Code

Families can agree on a private code word or phrase that is not shared publicly. If an unexpected emergency call arrives, ask the caller to provide the agreed code before considering any request for money.

Choose something difficult to guess and do not post it online. If the code may have been exposed, replace it. A code word is an additional safeguard, not a guarantee, so independently contact the person whenever possible.

3. Never Let Urgency Replace Verification

A caller may insist that someone will be harmed, arrested or left without help unless money is sent immediately. That pressure is a reason to slow down, not a reason to skip verification.

Before sending money, confirm the emergency through a separate, trusted communication channel. If the request involves a business payment, follow the organisation's normal approval procedures, even when the voice sounds like a senior employee.

4. Be Careful With Voice Recordings Online

Think carefully about the personal information and recordings you share publicly. Videos, interviews, voice notes and other audio can reveal material that an impersonator may attempt to misuse.

You do not need to stop speaking online or remove every recording of yourself. Instead, review your privacy settings, limit access to personal content where practical and be cautious about sending voice recordings or sensitive information to people you do not trust.

5. Protect Your Accounts and Verification Codes

Use strong, unique passwords and enable multi-factor authentication where available. Never share a one-time password, login approval or authentication code with someone who contacts you unexpectedly.

A criminal may use a convincing voice to persuade you to reveal a code or approve an account login. Once access is obtained, the attacker may attempt to impersonate you and contact people in your address book.

6. Verify Unexpected Payment Requests Through a Second Channel

If a friend asks for money, contact them directly through a previously established number or another trusted channel. For business transactions, verify new bank details or unusual transfer instructions with the appropriate person using an independent contact method.

Be especially cautious about requests for cryptocurrency, gift cards or transfers to unfamiliar accounts. These methods can make it difficult to recover money after it has been sent.

What to Do If You Receive a Suspicious Call

If you suspect a voice-cloning scam, do not argue with the caller or provide more personal information. End the conversation and verify the claim independently.

  • Do not send money or share account details.
  • Contact the real person using a trusted number or communication channel.
  • Save relevant call logs, messages, payment details and other evidence where safe to do so.
  • Block and report suspicious accounts or numbers through the relevant service.
  • Warn affected relatives, colleagues or contacts if the scammer may be impersonating them.

If you have already sent money, contact your bank or payment provider immediately and explain that the transaction may be fraudulent. Ask whether it can be stopped, recalled or investigated. Report the incident to the relevant law-enforcement or cybercrime reporting authority in your country.

In Nigeria, victims can contact their bank promptly and report suspected fraud to the appropriate law-enforcement authorities. Keep transaction references and communications that may assist an investigation. Recovery is not guaranteed, but acting quickly can be important.

The Future of Voice Cloning and Digital Trust

Voice cloning illustrates a broader challenge as artificial intelligence becomes more capable of producing convincing digital content. The same technology that can support accessibility, communication and creative work can also be misused to imitate people without their consent.

As generated audio improves, society may need stronger ways to verify identity in sensitive situations. Businesses could strengthen payment approval procedures, communication platforms could improve fraud detection, and technology developers could explore methods for authenticating the source of audio.

However, no single detection tool or watermark should be treated as a complete solution. Detection methods can make mistakes, and criminals may adapt their tactics. Verification procedures, security practices and public awareness will remain important.

The central lesson is simple: a person's voice is a familiar signal, but it is not a secure identity check. When money, private information or someone's safety is involved, trust should be supported by independent verification.

Key Facts About AI Voice Cloning Scams

  • AI can generate speech that resembles a real person's voice.
  • Criminals may use cloned voices in fake family emergencies, business fraud and impersonation schemes.
  • A familiar voice or displayed caller name does not prove identity.
  • Urgent demands for money, secrecy or authentication codes are warning signs.
  • Calling back through a trusted number is safer than relying on the original call.
  • Family verification phrases and workplace payment controls can add protection.
  • If money has been sent, contact the bank or payment provider immediately and report the suspected fraud.

Sources and Further Reading

For additional guidance, consult these official resources:

DAILY TOUCH INSIGHTS

Independent coverage of technology, business and the digital economy.

AI Voice Cloning Scams: How to Stay Safe