AI Scams Targeting Teenagers and Families

Quick answer

How has AI changed the scams reaching Australian families? Mainly by removing the old tells. A voice that sounds exactly like your teenager, a message written in flawless, familiar language, a profile photo that looks entirely real: AI can now produce all three cheaply and convincingly, which means the badly spelled email or the obviously fake caller is no longer the main threat. The defence that still works is not sharper listening, because a clone can sound perfect. It is a habit of verifying through a second channel before acting on anything urgent, and a simple family codeword that no scammer can guess. This guide covers both, plus what to teach your teenager about their own exposure.

How AI has changed the shape of a scam

A scam used to rely on volume and hope: send enough fake emails, and a few people will fall for the bad grammar and obvious pressure tactics. AI removes the need for volume to compensate for quality. A scammer can now generate a fluent, personalised message in seconds, clone a familiar voice from a short public clip, and build a profile photo that survives a casual glance, all with tools that require no special skill to use.

The result is not a new category of scam so much as the old ones - urgent calls, fake friend requests, phishing messages - arriving with far fewer warning signs. That is precisely why the defence has to shift from spotting mistakes to verifying identity, a theme that runs through the wider question of whether AI is safe for teenagers and the broader project set out in AI education for teenagers in Australia.

Voice cloning: the scam that sounds like your kid

The scenario that worries parents most, understandably, is a phone call that sounds exactly like their teenager, distressed and asking for help or money. Voice-cloning tools now need only a short sample of real audio, easily lifted from a public video, voice note or livestream, to produce a convincing imitation.

The uncomfortable truth is that your ear is not a reliable defence here. A clone built from a few seconds of real audio can be close enough to fool a parent under pressure, which is exactly the moment scammers are counting on. The protection is not listening harder. It is having a verification step that does not depend on how the voice sounds at all.

Phishing at scale and fake profiles

Beyond voice, AI lets scammers run convincing text-based attempts at real scale: messages that read like they came from a school, a bank, or a friend, each one personalised just enough to feel credible. Fake social media and gaming profiles built with AI-generated photos are a related risk that lands directly on teenagers, who spend real time on platforms where a stranger's profile can look entirely ordinary.

The tactic in both cases is the same: borrow trust from something familiar - a known brand, a known contact, a known platform - and pair it with urgency, so the target acts before checking. Teenagers are a frequent target directly, through gaming and social platforms, and a frequent unwitting entry point when scammers use their identity or accounts to reach parents and grandparents.

The family codeword: a simple defence that works

One habit does more work than almost anything else here: a family codeword. Agree, in person, on a word or short phrase that any real family member can supply if asked to prove who they are during an urgent call or message. Choose something that is not a birthday, a pet's name, or anything a stranger could find from a social media profile.

The rule that makes it work is equally simple: any urgent request for money, details or help that does not include the codeword when asked gets treated as suspicious, no exceptions, even if it sounds exactly right.

Verify-then-trust habits

Frame verify-then-trust as a life skill rather than a fear campaign. Urgency is the biggest giveaway of any scam, AI-powered or not: a message or call that demands a fast decision, before there is time to check, is behaving exactly like a scam behaves, regardless of how convincing it sounds or looks. Encourage your teenager to pause, verify through a second channel, and never feel embarrassed about double-checking, even with a friend. The families who fare best treat verification as completely normal, not as an accusation that something is wrong.

Common mistakes families make

• Trusting a voice or photo because it looks or sounds right, when both can now be convincingly faked.
• Reacting immediately to urgency, which is precisely the response a scam is engineered to produce.
• Never agreeing on a family codeword, leaving no fast way to verify a genuine emergency from a fake one.
• Assuming scams only target adults, when teenagers are frequently the entry point or the direct target.
• Treating one warning conversation as enough, rather than revisiting it as the tools and tactics change.

The recommendation: set a family codeword this week, agree on it in person, and make "verify through a second channel" the default response to anything urgent, no matter how convincing it sounds. That single habit outlasts every new scam technique AI makes possible, because it does not depend on spotting a fake. It depends on never acting on an unverified request in the first place.