For years, some of the easiest scams to spot were also the ones that couldn’t spell.
The unexpected email from your bank would have questionable grammar. The parcel notification would use slightly odd English. The logo might look stretched, the formatting would be wrong and the sender’s attempt to sound authoritative would instead read like someone had fed a corporate dictionary through a blender.
We learnt to look for those signs.
Check the spelling. Look at the sender. Hover over the link. Be suspicious of something that doesn’t quite look right.
That advice hasn’t suddenly become useless.
But the absence of those warning signs is no longer evidence that something is legitimate.
That’s an important distinction during Scams Awareness Week 2026, because generative AI has dramatically lowered the effort required to produce convincing written communication.
The spelling mistake is no longer your friend.
When the scam looks perfectly legitimate
Lynette Owens, Vice President of Consumer Education and Marketing at TrendLife, argues that some of the traditional advice Australians have relied upon is struggling to keep up.
“By and large, the old advice to look for misspellings and poor grammar is no longer enough,” Owens told Digital Reviews Network.
“AI has removed the hurdle they once had to clear to create highly persuasive and personalized scams. We should never assume well-produced content is NOT the work of scammers.“
That’s perhaps the most important part of the discussion.
A professional-looking email proves that someone can produce a professional-looking email.
It doesn’t prove they’re your bank.
A grammatically perfect message from your boss doesn’t prove your boss wrote it.
A convincing invoice doesn’t prove the bank account on it belongs to the company you’re expecting to pay.
And a phone call that sounds remarkably like somebody you know is increasingly something that needs to be considered in the context of voice cloning and other generative AI technologies.
Scammers haven’t stopped making mistakes. Poor spelling, strange URLs, mismatched domains and odd formatting remain perfectly good reasons to become suspicious.
The problem is the reverse assumption.
The absence of red flags isn’t a green flag.
Australians are still losing serious money
This isn’t a theoretical problem.
Scams Awareness Week runs from 24 to 28 August this year under the theme “No one’s just a number“, putting faces and stories behind Australia’s scam statistics.
Earlier figures from the National Anti-Scam Centre show the scale of the financial damage.
During the first three months of 2026 alone, Scamwatch received 45,816 reports, with reported losses of $76.7 million. ReportCyber separately recorded $187.7 million in reported scam losses during the same period.
After removing duplicates between the two reporting systems, the National Anti-Scam Centre put combined reported losses at $248.3 million in just three months.
There is some encouraging news in those numbers: reported losses were declining compared with the same period in 2025.
But $248 million remains an extraordinary amount of money.
Online contact is particularly significant. Scamwatch reports show Australians lost $38.3 million during the March quarter to scams beginning through online contact, with fake websites, advertising, social media and mobile apps accounting for around half of reported Scamwatch losses.
The National Anti-Scam Centre also took down 5,834 scam websites during those three months and referred 511 Facebook advertisements, profiles and groups to Meta for investigation.
You can read the National Anti-Scam Centre’s March-quarter 2026 figures here.
This is an industry operating at scale.
AI potentially makes scaling it easier.
Scammers don’t just attack technology. They attack moments.
Recent TrendLife research adds another dimension to the problem.
Its Digital Life and AI Experiences study surveyed 10,350 adults across nine countries between 25 March and 3 April 2026, including 1,038 Australians, with the Australian results weighted to be nationally representative by age, gender and household income.
Twenty per cent of Australian respondents reported either being scammed themselves, or knowing someone close to them who had been scammed, while making a major purchase or investment such as buying a car or property.
Scams were also reported around job interviews, government benefits and tax assistance, moving home, starting a new job and buying or selling property.
The full TrendLife research and methodology is available here.
There’s a common thread running through many of those situations.
Pressure.
Buying a house involves deadlines, deposits, agents, conveyancers, banks and communications from people you’ve possibly never dealt with before.
Buying a car can involve finance, insurance and transferring a substantial amount of money.
Starting a new job means communicating with unfamiliar people while legitimately being asked for personal information.
Scammers don’t necessarily have to manufacture an entire situation.
They can insert themselves into one.
And that’s where another traditional piece of scam advice — if it looks suspicious, stop — becomes inadequate.
What if it doesn’t look suspicious?
Stop judging the message and consider what it wants you to do
Owens says one useful indicator is not necessarily how a message looks, but the emotional response it is trying to create.
“If a text message, social media post, or email looks legitimate by every measure but is designed to raise your anxiety in any way – you’re late paying a bill, a fantastic discount will end soon, your credit card is having issues – this is a red flag and an opportunity to pause and research before hastily responding,” she said.
That’s a much more useful way to think about modern scams.
Instead of asking only:
Does this look like a scam?
Ask: What is this message trying to make me do?
Does it want money?
Does it want credentials?
Does it want a verification code?
Does it want personal information?
Does it want you to change bank details?
Does it want you to move a conversation to another platform?
And, critically: Does it want you to do it quickly?
Urgency is valuable to a scammer because time is the enemy of deception.
Give someone enough time to think, make a phone call, search independently or ask somebody else and the scam becomes much harder to sustain.
Don’t verify a message using information in the message
Owens recommends a minimum two-step process.
The first is independent verification.
“Check directly with the company or brand that’s trying to communicate to you and never trust a ‘push’ communication alone,” she said.
That distinction is critical.
If you receive an SMS supposedly from your bank telling you about suspicious transactions, don’t verify the SMS by clicking the link contained in the SMS.
Don’t call the number the suspicious message tells you to call.
Open the bank’s app yourself. Type its known website address yourself. Find the telephone number independently or use one you already know to be legitimate.
The same principle applies elsewhere.
If a supplier emails saying its banking details have changed, call your established contact.
If your conveyancer sends unexpected instructions for a house deposit, independently confirm them before moving the money.
If someone claiming to be a family member urgently needs money, establish their identity through another means.
If a recruiter unexpectedly approaches you and starts requesting sensitive information, independently establish that the company and recruiter exist and contact the organisation through details you sourced yourself.
You’re no longer trying to establish whether the message looks authentic.
You’re establishing whether the person behind it actually has the authority they claim to have.
There’s an important difference.
What about AI scam detection?
The second part of Owens’ recommendation is technological.
She suggests consumers use scam-protection tools, including TrendLife’s own ScamCheck, as another method of assessing suspicious communications.
“It may take a few more steps and moments of your time, but that cost is much lower than what is at risk if you respond without doing so,” Owens said.
There is merit in using technology as another layer of defence.
But there’s an important caveat.
A scam-detection tool shouldn’t simply become the new thing we trust blindly.
We specifically asked Owens about the limitations of anti-scam technology, what happens when technology gets something wrong, and how consumers can avoid transferring their trust from their own judgement to an automated detection system.
Her answer emphasised multiple layers rather than relying upon one.
“Again, it’s a minimum 2-step process: do your own verification and use a tool to help you. We would also recommend leaning on others you know and trust,” she said.
That last part is worth remembering.
Technology can be another signal. Another person can be another signal. Independent verification can provide another.
None needs to carry the entire burden.
Any detection system can potentially miss something malicious or flag something legitimate. More importantly, scammers change their techniques precisely because defenders learn to recognise the previous ones.
A green light from a security product therefore shouldn’t mean this is definitely safe any more than perfect grammar should mean this definitely came from my bank.
Use the technology.
Don’t outsource the decision to it.
Stop. Check. Protect.
Interestingly, that brings us remarkably close to the official Scams Awareness Week advice.
Scamwatch’s message is Stop. Check. Protect.
Stop before giving money or personal information.
Check that the person or organisation is really who they claim to be.
Protect yourself by acting quickly if something goes wrong and reporting scams to help protect others.
More information and official Scams Awareness Week resources are available through Scamwatch.
It’s simple advice, but perhaps the “check” part needs more emphasis as scams become more sophisticated.
Checking doesn’t mean rereading an email and deciding it looks convincing.
It doesn’t mean looking at the company logo.
It doesn’t mean deciding the person on the phone sounds trustworthy.
And increasingly it doesn’t mean looking for spelling mistakes.
It means going outside the communication you’ve received and independently establishing whether the request is genuine.
AI may be making scams more polished, personalised and convincing.
Our response shouldn’t be to become better at guessing.
It should be to make guessing unnecessary.
You don’t need to prove that something is a scam before refusing to follow its instructions.
Before handing over your money, credentials or personal information, make the person asking for them prove they’re legitimate.
Our thanks to Lynette Owens, Vice President of Consumer Education and Marketing at TrendLife, for taking the time to answer our follow-up questions for this article.
