AI-Powered Scams Cost Australians $2.18 Billion in 2025

ASIC dropped an uncomfortable update on 8 April. Australians lost $2.18 billion to scams in 2025 — up 7.8% on the year before, according to the National Anti-Scam Centre’s annual Targeting Scams report. Investment scams alone accounted for $837.7 million. And artificial intelligence, according to the regulator, is the reason the problem is accelerating.

For a business owner, this isn’t a consumer protection story you can scroll past. It’s an operational risk that’s already in your inbox.

ASIC removed 11,964 phishing and investment scam websites in 2025 — a 90% increase on the prior 12-month period, when it took down 6,270. That’s 32 sites a day. Commissioner Alan Kirkland was blunt: scammers are using AI to make fake investment ads “more polished, more convincing and harder to spot.”

The playbook has shifted. Where a scam email used to arrive with a broken logo and mangled grammar, today’s AI-generated version has perfect formatting, tailored language, and sometimes a deepfake video of a public figure endorsing a trading platform. ASIC pulled more than 1,100 investment scam ads from social media in 2025. Some featured AI-generated videos claiming trading bots could predict stock prices with 85% accuracy and deliver 374% returns.

Investment scams get the headlines, but payment redirection scams — the kind that hit businesses directly — accounted for $166.8 million in losses in 2025, up 9.3% year on year, per the NASC report. This is the one that should keep you up at night. A scammer intercepts or fabricates an invoice from a known supplier, changes the bank details, and sends it to your accounts team. AI makes every part of that fraud more convincing — the email tone, the invoice layout, even a follow-up phone call using cloned voice audio.

A CommBank study of nearly 2,000 Australians, published in January 2026, found that 89% are confident they can spot an AI scam. In testing, only 42% actually could — worse than a coin flip. Among small business owners, only 55% had cross-checked supplier payment details in the previous six months. Half of all deepfake scam attempts arrived by email — the same channel your quotes, invoices and supplier communications travel through every day.

If you run a trades business, your exposure is on the invoice side. You send dozens of quotes and invoices a week. Your customers receive them by email. Each one is an opportunity for a scammer to intercept, duplicate and redirect. Small businesses with lean admin teams and limited verification processes are disproportionately vulnerable — you don’t have a dedicated fraud department, and the person paying supplier invoices is probably the same person answering phones.

If you run an accounting or financial advisory practice, the risk is doubled. Your clients are targets for investment scams — the $837.7 million category. They’re seeing AI-generated ads on social media promising returns no legitimate investment delivers. When they lose money, they’ll ask why nobody flagged it. ASIC’s Moneysmart website published new AI guidance on 23 March specifically because younger Australians are increasingly turning to AI tools for financial advice, often without verifying the source. We covered that trend in an earlier note.

First, verify payment details independently. Any invoice above a threshold you set — $5,000, $10,000, whatever fits your operation — gets a phone call to a known number before payment is processed. Not the number on the invoice. A number you already have on file. This single step eliminates most payment redirection fraud.

Second, if you’re in professional services, talk to your clients about AI scams. A two-minute conversation about deepfake investment ads is worth more than a quarterly compliance newsletter. ASIC’s Moneysmart site has consumer-facing resources you can share directly.

Third, treat AI literacy as a business risk control. The businesses that understand how AI generates text, images and video are the ones that recognise when something doesn’t add up. This isn’t about adopting AI for productivity — though that matters too. It’s about knowing enough to protect what you’ve already built.

Key takeaways

Want to understand where AI risk sits in your operations?

A short conversation can map the exposure points — from invoice verification to client-facing communications. Book a call to talk it through.