Deepfake Detection and Response for Perth Businesses

A handful of seconds of audio from a public video is now enough to clone a voice convincingly. For businesses, that's turned what used to be science fiction into a practical scam vector - and one that doesn't trip a single piece of email security software, because there's no email or malware involved at all.

How Deepfake Scams Actually Target Businesses

• Voice cloning for urgent payment requests - a call that sounds exactly like a director or manager, asking for an urgent transfer or gift card purchase, often with a plausible reason for secrecy or urgency.
• Video deepfakes in video calls - increasingly used to impersonate executives in video meetings to authorise payments or share sensitive information, exploiting the trust placed in "seeing" someone on a call.
• Synthetic identity documents - AI-generated or manipulated ID documents used to pass verification checks during onboarding or account setup.

Warning Signs Worth Knowing

Detection is getting harder as the technology improves, but a few patterns still hold up:

• Unusual urgency combined with a request to bypass normal process - this is the single biggest tell, regardless of how convincing the voice or video is.
• A channel switch mid-conversation - starting on a call and being asked to move to email or text to "confirm" details, often to avoid further real-time scrutiny.
• Audio or video artefacts - unnatural pauses, slightly off lip-sync, inconsistent lighting, or a voice that sounds right but the phrasing feels subtly wrong.
• A request that's out of character - if the person being impersonated wouldn't normally make this kind of request by phone or video, that mismatch matters more than how convincing the voice sounds.

Building a Response Process

Detection alone isn't a reliable defence - the technology will keep improving. A documented response process is what actually limits the damage:

1. Set a verification rule for anything involving money or sensitive data - any payment or detail change requested by call or video must be confirmed through a separate, pre-established channel before action is taken. This single rule defeats the vast majority of deepfake scams regardless of how convincing they are.
2. Establish a code word or shared verification method for genuinely urgent, out-of-process requests between senior staff - a low-tech but effective control.
3. Make it normal to say no to urgency - staff need explicit permission to delay and verify, even when a request appears to come from a senior figure, without fear of seeming obstructive.
4. Know where to report it - to your bank immediately if money has moved, and to ReportCyber for a record and guidance.

Where to Start

Most businesses don't need deepfake-detection software - they need a verification habit that doesn't depend on spotting the fake at all. That makes this less a technology problem to solve and more a process to put in writing and train staff on, alongside your existing approach to invoice fraud and Business Email Compromise.

Frequently Asked Questions