Is This a Scam?
Ask an AI.
A second pair of eyes lives in your pocket now. Let's practise using it.
Two real-world scenarios: the parcel email that isn't from Australia Post, and the voicemail that isn't from your grandson.
Scams arrive when you're busy,
and they're built to rush you
You often feel something is off. What's been missing is a quick, judgement-free way to check that feeling.
Copy it. Paste it. Ask:
"Does this look like a scam?"
An email says your
parcel is on hold
You're expecting a delivery, as most of us are, most of the time. This lands in your inbox. It has the logo, the colours, the tracking number.
You're just not sure. Something feels a little off, but you can't say what.
To: you
Dear Customer,
Your parcel AP-7729-041-NT could not be delivered and is being held at our depot. To avoid it being returned to sender, you must confirm your address and pay a redelivery fee of $2.99 within 24 hours.
Confirm details & pay fee
https://auspost.redirect-parcel.info/track?id=7729041
Australia Post Customer Services. If you do not action this notice your item will be returned.
Four flags, one habit
"Hey Gran,
I lost my phone"
A voicemail from a number you don't know. Your phone has written out the words for you. It sounds exactly like something your grandson would say.
Your chest tightens. He needs help. And that feeling is exactly what the message is designed to produce.
It has read every warning
about this exact scam
The "Hi Mum" or family impersonation scam is one of the most heavily documented scams in Australia. The ACCC issued a national warning after more than 1,150 Australians lost $2.6 million to it in the first seven months of 2022 alone, and it has kept circulating ever since; women over 55 made up two thirds of victims.
Why scam-checking is a job
AI is actually good at
Pattern recognition
Scam formats repeat worldwide and are described in thousands of public warnings. Matching your message against those patterns is a low-risk, high-skill task for an AI.
It explains itself
You don't just get "scam" or "safe". You get the reasons; the odd sender address, the pressure tactics, the link that doesn't match. The reasons teach you.
Hallucination risk is low
AI systems can invent facts when asked to recall specifics. Here they aren't recalling; they're reading text you gave them. That's the kind of task where they are most reliable.
A good question gets
a useful answer
A prompt that works
"I received this message and I'm not sure about it. Does it look like a scam? What are the red flags, and what should I do next?"
[paste the message]
Leave your details out
The scammer's message is not private; your details are. Remove or blank out your own name, address, account numbers or codes before pasting. The AI doesn't need them to judge the message.
You get good at AI the same way
you get good at anything: practice
The people who benefit most from AI assistants are the ones who have practised talking to them; who know what they're strong at, where they wobble, and how to ask a better question. Scam-checking is the perfect practice ground: real stakes, clear answers, and every session teaches you the red flags yourself.
Stop. Paste. Verify. Report.
Where today's numbers come from
The two scenario messages are fictional composites built from the patterns documented in the sources above. The simulated AI replies are realistic examples written for this session, not live output.