How to spot AI and deepfake impersonation scams before they catch you out

You receive a call from a family member asking for urgent financial help. It's an unknown number, but their voice sounds identical. What do you do?

Or maybe you see a video of a celebrity sharing what looks like an amazing investment opportunity. Do you click the link?

A few years ago, scams that look this real might have seemed unlikely. Today, advances in artificial intelligence (AI) mean scammers can create convincing fake voices, videos and messages in a matter of minutes.

It can feel pretty scary. Research from the Australian Institute of Criminology found around half of Australian adults worry about becoming victims of AI-enabled crime, including impersonation, deception and deepfake content. And there's reason to be concerned. In the 2025-26 financial year the Australian Securities and Investments Commission (ASIC) reported Australians lost more than $7.4 million on online impersonation scams. 

But arming yourself with the knowledge of how these scams work can make them much easier to spot, and make you much less susceptible to falling for them.

What are AI and deepfake scams?

AI scams use artificial intelligence to make scams more convincing, personalised and harder to detect.

A deepfake is a type of AI-generated content that uses technology to create realistic fake videos, images or audio recordings that appear genuine. Sometimes a deepfake may appear to show a celebrity, sometimes it might be your own friends and family.

What makes these scams different to regular phishing scams is that they can appear highly believable. A scammer may be able to:

• Clone a person's voice from a short audio clip

• Create a fake video of someone appearing to say something they never said

• Generate personalised messages that reference details about your life

• Build fake websites and documents that closely resemble legitimate businesses

What types of AI and deepfake scams are common?

Fake bank, Government or business impersonation scams

Scammers may use AI-generated messages, emails or voice calls while pretending to represent a trusted organisation.

For example, you might receive a text claiming there's a problem with your bank account, followed by a phone call from someone who appears professional and informed.

The goal is often to create urgency and convince you to share personal information, passwords or verification codes.

Voice cloning scams

These days, scammers can create a convincing voice clone using only a small sample of someone's voice. 

They may then contact family members pretending to be a loved one in distress and needing money urgently.

Because the voice sounds familiar, people can feel pressured to act quickly without stopping to verify the request.

Celebrity investment scams

A common tactic involves deepfake videos of well-known public figures appearing to endorse an investment opportunity.

These advertisements often promise unrealistic returns, limited-time opportunities or guaranteed profits.

ASIC reported that the most commonly impersonated famous people by online scammers in 2026 was Anthony Albanese, Tom Piotrowski, Dick Smith, Gina Rinehart, Pauline Hanson and John Laws.

Remember, seeing a familiar face doesn't mean the opportunity is legitimate. Deepfake technology can make fake endorsements look surprisingly real. 

Fake websites and documents

AI tools allow scammers to rapidly create professional-looking websites, identification documents and certificates. Some fake websites closely mimic the branding, logos and appearance of legitimate organisations. 

That's why it's important to type website addresses directly into your browser rather than clicking links from unsolicited emails, text messages or social media advertisements.

What are the warning signs of an AI-powered scam?

While the technology behind scams is evolving, the tactics often stay the same.

Watch out for:

• Unexpected contact from someone claiming to be from a bank, Government agency or business

• Requests for personal information, passwords or security codes

• Pressure to act immediately or keep something secret

• Investment opportunities promising guaranteed returns

• Messages that create panic, excitement or urgency

• Requests for money transfers, gift cards or cryptocurrency

• Communication from a new phone number claiming to be someone you know

Most importantly, don't assume a voice recording, video or image is proof that someone is who they claim to be – it's best to always double check via a separate channel.

How can you protect yourself from AI and deepfake scams?

Slow down before you act

Scammers rely on people making quick decisions. If you're asked to transfer money, share personal information or click a link, pause and take a moment to assess the situation. A few extra minutes could make all the difference.

Verify requests through another channel

If a family member, friend or organisation contacts you unexpectedly, check with them on another platform. Call the person back on a known number. Visit the organisation's official website and use their published contact details. You can also jump onto the ASIC website to confirm if the person offering the investment is licensed. Don't rely on the details provided in the message itself.

Be cautious about what you share online

Many scammers gather information from social media and public websites to make their approaches more convincing. Review your privacy settings and think twice about which personal details you share publicly.

Never share authentication codes

Banks and legitimate organisations will not ask you to share one-time passcodes, internet banking login details or security credentials. Treat these details like the keys to your home – keep them private.

Trust your instincts

If something feels unusual, inconsistent or too good to be true, it's worth investigating further. It's always better to double-check than rush into a decision you'll later regret.

What should you do if you think you've been targeted?

If you think you've encountered an AI scam, act quickly to limit the damage:

• Stop all communication with the scammer

• Don't send money or provide additional information
  

• Contact your bank immediately if you suspect a scam or if you've shared banking details or made a payment

• Change passwords for any affected accounts

• Report the scam to Scamwatch

• Reach out to IDCARE, Australia and New Zealand’s national identity and cyber support service, for free advice and to help you make a plan to limit any damage if your identification details have been accessed.  

If you believe your AMP Bank account or superannuation has been compromised, alert us immediately. 

AMP Bank customers can call 13 30 30. 

AMP Bank GO customers can message us 24/7 in the app, or call us on 1800 950 105. 

AMP Super members can call 13 12 67.

Read more about phishing scams, investment scams and romance scams here.