Fake Nvidia keynote deepfake fooled 100,000 viewers on YouTube

Nearly 100,000 people watched a fabricated livestream believing it was real
YouTube’s algorithm boosted a fraudulent event over Nvidia’s verified broadcast
Viewers were invited to scan a QR code for a crypto giveaway

In a sharp reminder of how easily digital manipulation can spread, a fake Nvidia GTC keynote featuring an AI-generated version of Jensen Huang attracted close to 100,000 viewers on YouTube.

The fraudulent live stream promoted a supposed “crypto mass adoption event,” and it was hosted by a random channel named Offxbeatz.

Dylan Martin, the real Nvidia GTC DC keynote had just begun when the fake broadcast rose to the top of YouTube’s search results.

By the time Huang began speaking at the legitimate event, only about 12,000 people were watching, compared to nearly 95,000 following the fabricated stream.

The AI-generated “Jensen” greeted the audience and promised an exclusive announcement about cryptocurrency that aligned with Nvidia’s supposed “mission to accelerate human progress.”

What followed was a scripted narrative praising Nvidia GPUs for powering blockchain networks and digital payments.

Also Read Amazon has a massive holiday sale with Black Friday favorites – 60% off Ninja, Apple, Nespresso, and Shark

The entire setup ended with a QR code inviting viewers to participate in a crypto distribution scheme, a textbook example of an online financial scam.

Despite the slick presentation, several signs made the deception fairly obvious.

The deepfake’s speech patterns were slightly unnatural, and the exaggerated claims about cryptocurrency adoption should have raised suspicion among careful viewers.

Yet tens of thousands of people kept watching, showing how easily strong visuals can overpower basic skepticism.

YouTube eventually removed the stream, but its brief success shows how algorithmic promotion can elevate fraudulent content above verified sources.

The video, described as a product of modern AI trickery, was convincing enough to mislead thousands of viewers searching for the event.

It shows how deepfake technology has outpaced many platforms’ current moderation systems.

Without stronger verification methods, such synthetic impersonations will likely continue to spread.

Traditional cybersecurity measures, such as antivirus software and firewalls, remain vital for device-level protection.

Unfortunately, they offer little defense against large-scale manipulation that relies on social engineering and real-time deception.

Therefore, platforms must invest in better identity verification tools that can authenticate live streams before they go public.

Beyond platform responsibility, users also need stronger skepticism when engaging with online events, especially those involving cryptocurrency transactions.