Security News Deepfakes Make Video Calls Dangerous—On-Device Detection Is The Answer

[Brownie2019]

Content source

https://www.forbes.com/sites/larsdaniel/2026/06/02/deepfakes-make-video-calls-dangerous-on-device-detection-is-the-answer/

For years, the video call solved a basic trust problem. If an email looked suspicious, a person could get on camera and confirm the request. In business, family disputes, online dating, remote work and even legal settings, seeing and hearing someone became a practical shortcut for believing them.

Deepfakes are breaking that shortcut. The old assumption that seeing someone on screen is enough is no longer safe. I have written before about why seeing is believing is dead for visual evidence. Live video calls bring that same problem into the moment when people are making decisions.

The Arup case shows the problem. In its 2024 financial statement, Arup said criminals used “fake voice, signatures and images” to execute a fraud, while also stating that its networks were not compromised and that no personal or project information was accessed. Arup’s financial statement described the incident as a social engineering-led attack, not a traditional network breach.

Read more:

Deepfakes Make Video Calls Dangerous—On-Device Detection Is The Answer

Why catching deepfake video calls belongs on your device—not in the cloud.

www.forbes.com

 

[Bot]

Read more:

Deepfakes Make Video Calls Dangerous—On-Device Detection Is The Answer

Why catching deepfake video calls belongs on your device—not in the cloud.

www.forbes.com

That risk is real, but I would be careful with one part of the framing: video calls are not “dangerous” by themselves. The bigger issue is that video and audio can no longer be treated as strong identity proof on their own.

What has changed

Deepfakes lower the reliability of cues people used to trust:

• Face and voice recognition during live calls
• Urgency-based requests from “executives,” relatives, or partners
• Requests to bypass normal approval steps because “you can see it’s me”

That makes live video less useful as a standalone verification method, especially for financial approvals, account recovery, and sensitive internal requests.

What the Arup example actually suggests

Based on the description you quoted, this appears to be a social engineering case using impersonation elements, not evidence that all video calls are unsafe.

The important takeaway is:

• A convincing call should not override process
• Identity verification needs independent checks
• Human trust shortcuts are now a major attack surface

Practical defense that works today

For most organizations, the best response is not to rely on any single “deepfake detector.” It is to combine process controls with technical safeguards.

• Use out-of-band verification for money transfers, banking changes, and sensitive approvals
• Require two-person approval for high-risk transactions
• Use known contact methods, not the contact details provided in the suspicious request
• Treat unusual urgency or secrecy as a red flag even if the person looks and sounds real
• Train staff that video presence is not equivalent to identity proof
• Keep records of approval workflows so exceptions are visible

About on-device detection

On-device detection may become one useful layer, but I would not call it “the answer” yet unless there is strong independent evidence behind a specific product or method. Detection tools can help, but attackers adapt quickly, and false positives or missed detections are both possible.

In general, stronger process design is more dependable than assuming AI can always detect AI.

Bottom line

Seeing someone on a live call is no longer enough for high-trust decisions. Video can still be useful for communication, but it should not be treated as proof of identity without independent verification.