Skip to content

Wednesday, November 18, 2026 Updated

EditionEnglishEspañol

TechnologyNews

Coinbase Uses AI to Stop $4.7M in Elon Musk Deepfake Scams

Coinbase stopped $4.7 million in scam attempts using an AI system that detected fake broadcasts featuring Elon Musk deepfakes.

By

2 min read

Coinbase IA estafas deepfakes a Elon Musk

Key points

  • $4.7 million prevention: Cryptocurrency platform Coinbase reported that its artificial intelligence-based fraud detection system prevented $4.7 million from reaching scammers during a recent event.
  • 91% transfer block: According to the official announcement released this Tuesday by the exchange, the automated defense mechanism successfully stopped approximately 91% of transfer attempts linked to the fraud.
  • Massive streaming reach: The fraudulent broadcast attracted more than 500,000 viewers on YouTube by mimicking an official Tesla event related to a flying vehicle demonstration.

Digital identity impersonation has become a major security challenge for the crypto industry. On Tuesday, November 17, Coinbase (NASDAQ: $COIN ) reported that its AI-powered fraud detection system successfully stopped $4.7 million in transfer attempts directed by malicious actors.

The exchange reported that the criminal scheme used an artificial intelligence-generated video simulating the image of Elon Musk to promote an alleged Tesla (NASDAQ: $TSLA ) cryptocurrency giveaway or distribution. The operators behind the scam recreated a live stream that imitated the official broadcast of a flying vehicle demonstration held the previous month.

The fraudulent broadcast incorporated visual elements designed to capture audience attention, including an Elon Musk deepfake, specific digital wallet addresses, and a QR code. The projections promised viewers that any digital asset sent to the indicated addresses would be doubled and returned immediately.

Coinbase IA estafas deepfakes a Elon Musk

Intervention Mechanism and Automated Response

Coinbase’s report details that, over the course of about two hours, hundreds of clients attempted to transfer multi-million-dollar sums toward the addresses shown in the broadcast. The company clarified the nature of the transactions by stating verbatim: “Funds moved because the customer was convinced, not because an attacker forced their way in.”

Facing this situation, Coinbase’s security systems identified suspicious movements by analyzing patterns associated with fake giveaway schemes, brand impersonation, and the recent creation of destination addresses. Once a transaction triggered alerts, an artificial intelligence agent intervened in an automated manner.

The software operated by explaining how this type of scam works and directly questioning users about the nature of the offer before transactions were concluded. According to institutional documentation, this defense model did not require human intervention and operated continuously from the first detected transfer to the last, despite the fact that this specific form of deception was not part of its initial prior training.

The company added that the system recognized the typical structure of the fraud, characterized by the promise of multiplying funds, the urgency generated by a short-duration event, and the irreversibility of digital transfers. Similar incidents have previously affected audiences on video platforms using Elon Musk‘s image during corporate events by SpaceX (NASDAQ: $SPCX ).

About the author

C. Monasterio

Passionate journalist specialized in cryptocurrency, blockchain technology, and decentralized finance (DeFi). With years of experience covering market trends, project analyses, and global crypto regulations, she brings accurate, insightful, and engaging content to readers worldwide. Her work focuses on making complex blockchain topics accessible to everyone while maintaining high standards of credibility and journalistic integrity.

Follow Crypto Economy on Google News