← Field Notes
·19 August 2026·4 min read

AI Deepfakes Now Impersonate Australian Businesses

ASIC removed 19,400 scams in a year, up 182%, as AI deepfakes clone Australian businesses and bosses. Why 'looks right' no longer verifies — and the fix.

ASIC removed a record 19,400 scams from the internet in the year to June 2026 — up 182 per cent on the year before — and the reason they keep multiplying is that AI now builds them. Deepfake videos, cloned voices, fabricated reviews and fake news articles, increasingly wearing the face and name of a real, identifiable Australian. Reports tied to that impersonation cost $7.4 million. The uncomfortable shift for a business owner: your own brand is now raw material for the fraud.

On 17 August, ASIC warned that scammers are using generative AI to spin what it called 'vast webs of deception' — linked networks of fake endorsements, websites and reviews built to look independently verified. The most-impersonated names were public figures: Anthony Albanese, Jacqui Lambie, Angus Taylor, and finance commentators Tom Piotrowski and Alan Kohler. But the same toolkit reaches down to smaller targets — ASIC noted scammers also reuse legitimate Australian Financial Services licence details and impersonate real businesses.

ASIC Chair Sarah Court put the core problem plainly: 'AI is making investment scams more convincing and harder to detect. A simple online search is not enough to verify whether an opportunity is legitimate.' That sentence is the whole story. The check most people fall back on — does this look real? — has stopped working.

19,400

Scams ASIC removed in a year

Year to June 2026 — up 182% on the year before

$7.4m

Reported losses from impersonation scams

Scamwatch reports, per ASIC

244%

Rise in digital document forgeries

Year on year, per Moore Australia

Want to know what this means for your business? We'll map it in one call.

Book a call →

There are two ways this lands on a small business. The first is the one we've covered before — you get defrauded. AI writes a flawless invoice, clones a supplier's voice, and a payment walks out the door; we wrote about that, and the free rule that stops it. The second is newer and harder to shrug off: your business becomes the disguise. A cloned website, a fake review page, a 'you' on a video call — used to defraud your customers, your staff, or a supplier who trusts your name.

The barrier to entry has collapsed. A LinkedIn video, a voicemail greeting, or a few seconds of audio pulled from a webinar is now enough to build a convincing fake — what once took a skilled forger a week now takes minutes. In one case reported by SmartCompany, Noosa Council paid more than $2 million to a fraudulent account after scammers impersonated a contractor: a believable email, then a follow-up call to 'reassure' staff, then a request to change the bank details. The control most owners still lean on — does this look and sound right? — no longer holds, because the look and the sound are exactly what AI now fabricates. If you're not sure which of your own approvals a convincing fake could push through, a short call will find them.

For professional services firms — advisers, accountants, mortgage brokers — the exposure runs through identity: a licence number reused on a fake site, a client impersonated to move funds, a deepfaked partner authorising a transfer. For trades, it's the storefront: cloned business listings and fake reviews that quietly divert your customers, or a caller who sounds like the boss approving a subbie's 'updated' bank details. And it reaches your paperwork. AI-forged documents — payslips, licences, insurance certificates, invoices — now pass a visual check; Moore Australia reports digital document forgeries rose 244 per cent year on year. That matters every time you verify a new subcontractor, onboard a client, or make a hire.

Treat 'it looks right' as worthless, and build verification around a channel the scammer can't control. For any request to move money or change details, call back on a number you already had — never the one in the message — and add a second approver above a threshold you set. Agree a simple code word with your team for urgent money or identity requests; it's the one thing a deepfake can't reproduce. Verify documents against their source — a licence against the state register, a reference by calling the organisation directly — not by eye.

Then defend the identity you own. Claim and lock your Google Business Profile, watch for cloned sites and fake reviews, and tell customers your real contact channels so a fake one stands out. Report impersonations to ASIC and Scamwatch — those 19,400 takedowns started as reports — and, if you hold a licence, register your web addresses in ASIC's Professional Registers so customers can check them. None of this costs more than an afternoon. The alternative — as with the AI scams already costing Australians billions — is finding out the hard way that a familiar face is no longer proof of anything.

Key takeaways

ASIC removed a record 19,400 scams in the year to June 2026, up 182 per cent, with reported impersonation-scam losses of $7.4 million (ASIC media release, 17 August 2026).
Generative AI now builds deepfake videos, cloned voices, fake reviews and cloned websites — increasingly impersonating real Australian businesses and the people who run them.
Verification by appearance is dead: video and voice are both fakeable, so ASIC warns a simple online check is no longer enough.
Defend both directions — verify money and identity on a channel you control, and actively guard the brand, listings and documents scammers can clone.

Common questions

Can you still trust a video call to verify who you're dealing with?

No. Deepfake video and cloned voice are now good enough that ASIC says a simple check is not enough. Verify anything involving money or identity on a separate channel you control — a callback to a number you already had, or a pre-agreed code word.

What do I do if a scammer is impersonating my business?

Report it to ASIC and Scamwatch so it can be taken down — ASIC removed 19,400 scams in the year to June 2026, largely from reports. Claim and secure your Google Business Profile, and tell customers your real contact channels so a fake one stands out.

How can AI-forged documents affect my business?

AI can now produce payslips, licences and invoices that pass a visual check — digital document forgeries rose 244 per cent year on year, per Moore Australia — so verify credentials against their source register or issuer, not by eye, when hiring, onboarding or paying a new supplier.

Sources

ASIC — Scammers are using AI to spin vast webs of deception (26-195MR, 17 August 2026)

SmartCompany — From deepfake bosses to perfect emails: Scams SMEs must watch for in 2026

Moore Australia — Fraud is getting smarter: what business owners need to know about AI-enabled financial crime

Assumptions & methodology
  1. The 19,400 scams removed, the 182 per cent year-on-year increase, and the $7.4 million in reported impersonation-scam losses are from ASIC media release 26-195MR (17 August 2026). ASIC's takedown figures cover the financial year to 30 June 2026.
  2. The Noosa Council example and the LinkedIn-video / voicemail capability description are as reported by SmartCompany (2026); the $2 million loss figure is theirs.
  3. The 244 per cent rise in digital document forgeries is as reported by Moore Australia and reflects their cited industry data — it is not an ASIC or Australian-government figure.

Related notes

ShareLinkedInEmail

Field Notes are general commentary on AI trends for Australian businesses. They don’t constitute professional advice. Talk to your accountant, lawyer, or IT adviser before acting on anything specific to your situation — or talk to us if you want help working out where AI fits.

Don't miss the next one

Get each new Field Note in your inbox as it publishes — short, practical AI intelligence for business owners.

Could a convincing fake push a payment through your business?

A short call maps where money and identity decisions are made in your business — and where a verification step belongs. Book a call to talk it through.

Book a call →