Netarx

Real-time deepfake detection for voice, video and email. Shared awareness for enterprise communication.

[Netarx @netarxllc]

$5,345,846,501 confirmed stolen. 1.8B records exposed. 276 incidents, 584 primary sources.

Almost none of it started with a broken server. It started with someone picking up the phone.

The Global Social Engineering Impact Database is free and public: https://t.co/yCGbiEL1R2

[Netarx @netarxllc]

Your voice is on a webinar. Your face is on your team page. That's enough.

Clone yourself in under a minute and see what your team would be up against → https://t.co/uSoxoKNZGC

[Netarx @netarxllc]

Social engineering data lives in a thousand scattered one-off reports. So we built a database and made it public.

276 source-linked incidents $5.3B documented stolen 1.8B records affected 13% AI-involved

Free. No signup.
https://t.co/yCGbiEL1R2

[Netarx @netarxllc]

7/ The clone is free. Media purged in 24 hours, never used for training. Only you can clone you.
https://t.co/uSoxoKOxwa
#DataPrivacy #AIsafety #Netarx

[Netarx @netarxllc]

6/ That gap is what NIK closes. Real-time deepfake detection inside Zoom, Teams, email, voice, and messaging. Green, yellow, red, in the workflow, no integration.
https://t.co/uSoxoKOxwa
#NIK #DeepfakeDetection #ZeroTrust #Netarx

[Netarx @netarxllc]

5/ So the honest test is: clone yourself, put it in front of the people who move money, and see who bites. You'd rather learn it from your own clone than from someone else's.
https://t.co/uSoxoKNZGC
#RedTeam #SecurityTesting #FraudPrevention

[Netarx @netarxllc]

4/ Awareness training tells people to "look for the signs." There aren't reliable signs anymore. You can't train a human eye past a model that renders faster than the eye samples.
https://t.co/uSoxoKNZGC
#SecurityAwareness #AIThreats #CyberRisk

[Netarx @netarxllc]

3/ Arup lost $25M to a video call where every participant except the victim was synthetic. The finance employee did everything right by the old playbook. The old playbook is the problem.
https://t.co/uSoxoKNZGC
#CFOFraud #WireFraud #FinancialCrime #DeepfakeScam

[Netarx @netarxllc]

2/ The uncomfortable part isn't that it works. It's how little it costs. One photo. A few seconds of audio. No skill required. That's the whole attack surface now.
https://t.co/uSoxoKNZGC
#Deepfakes #SocialEngineering #InfoSec

[Netarx @netarxllc]

1/ We'll build a deepfake of you, free, from one photo and a few seconds of your voice.
Then you send it to your own team and find out what happens.
https://t.co/uSoxoKNZGC
#DeepfakeDetection #CyberSecurity #AIThreats

[Netarx @netarxllc]

Clone yourself. Send it to your team. See who bites.

Free deepfake of you from one photo + a few seconds of voice. ~2 min.

Then put it in front of whoever would approve a wire if you asked.

Most teams find out they'd have paid.

https://t.co/uSoxoKNZGC

[Netarx @netarxllc]

Awareness training taught your employees to stop and inspect every email.

That hesitation costs 7 hours per employee per year.

You have been paying for it since the training worked.

#phising #cybersecurity

[Defrag @Defrag_Avatar]

Upload a photo. Add a voice clip. Join the meeting as someone else in under 2 minutes.

That's the attack. Now it's your training exercise.

https://t.co/1SVoBD4YQq

[Netarx @netarxllc]

New launch: Advanced AI Email Models.

Content filters catch known malware and bad links. They miss a perfectly written email from a name you know.

We check the message against the sender's real relationships via the NIK social graph.
https://t.co/lvKfJZWa7z

[Netarx @netarxllc]

5/ This pattern shows up constantly. We catalogue it at https://t.co/yCGbiEL1R2, a free source-linked database of breaches and fraud that started with someone being deceived. Updated daily.
#cybersecurity #deepfake

[Netarx @netarxllc]

4/ The lesson isn't "better detectors."

We keep checking whether a face is real instead of whether the operation behind it is. A Fiverr account beat millions in detection R&D.

[Netarx @netarxllc]

3/ It worked. The network out-engaged the YouTube channels of the Washington Post and the New Yorker. 90,000+ real commenters.

Enforcement didn't come from a classifier. It came from Semafor and Riddance AI publishing.

[Netarx @netarxllc]

2/ The face.

They paid gig actors $26 a video off Fiverr and Backstage to read LLM-written scripts on camera. 24 hour turnaround, actor supplies their own lighting.

Every synthetic media detector YouTube has built saw a real human and moved on.

[Netarx @netarxllc]

Thread:
1/ YouTube just terminated 20 channels in a political video network that racked up 45 million views on AI-generated scripts.

The scripts, thumbnails, headshots and websites were all synthetic. One thing wasn't.
#Cybersecurity

[Netarx @netarxllc]

Applications to US remote IT roles carrying North Korean fraud patterns: 11% in Q3 2024, an estimated 47% last quarter, per Endorsed data via Fortune. Patterns, not confirmed operatives, but the slope is the story. https://t.co/yCGbiEL1R2