Skip to content
Technology & InnovationLatest

Coinbase AI Foils Musk Crypto Scam, Blocks Millions in Stolen Funds

Published

Coinbase announced that its artificial intelligence-powered scam detection system successfully intercepted $4.7 million, representing approximately 91% of funds targeted by scammers, during a fraudulent Elon Musk cryptocurrency giveaway.

MissedBlock Desk · 3 min read

Coinbase AI Foils Musk Crypto Scam, Blocks Millions in Stolen Funds

Coinbase AI Blocks Millions in Fake Elon Musk Crypto Scam

Coinbase announced that its artificial intelligence-powered scam detection system successfully intercepted $4.7 million, representing approximately 91% of funds targeted by scammers, during a fraudulent Elon Musk cryptocurrency giveaway.

What sets this incident apart is that the entire operation was managed without human intervention. The company’s AI agent identified the scam and intervened as customers attempted to send their money.

The scheme centered on a deceptive YouTube livestream that mimicked a Tesla flying car demonstration. This fake broadcast featured an AI-generated likeness of Elon Musk, displayed cryptocurrency wallet addresses and a QR code, and promised viewers double their investment if they sent cryptocurrency. The stream garnered over 500,000 viewers, providing the fraudulent event with significant exposure. According to Coinbase, a key challenge was that the scammers did not need to breach any accounts or steal login credentials.

Customers were already logged in, had access to their funds, and willingly authorized the transfers, believing the giveaway to be legitimate. This modus operandi makes the fraud particularly difficult to detect using traditional security measures, which typically focus on unauthorized account access or suspicious activity.

To counter this threat, Coinbase deployed an AI scam-prevention agent designed to flag risky transfers and challenge customers before their funds leave the platform. The system utilizes machine learning models to assess transactions in real-time, looking for warning signs such as giveaway-related activity, brand impersonation, and transfers to newly created wallet addresses.

When a payment is flagged, the AI agent engages customers in a targeted conversation, explaining the specific scam pattern and why the offer should be viewed with suspicion, rather than issuing a generic security warning. Coinbase also stated that each response generated by the agent is assigned a confidence score before being presented to the customer, helping the system determine when it has sufficient information to intervene decisively.

Notably, the AI agent had not previously encountered this specific scam. Deepfake giveaways linked to live product launches were not part of its training data. However, it still recognized familiar red flags, including promises of amplified crypto holdings, time-sensitive pressure tactics, and irreversible transfers. The system reportedly managed the flagged attempts from the initial transaction to the last without requiring human oversight.

“Previously, staging a scam like this required a crew, a studio, and weeks of work. Now, it takes a livestream slot and a model,” a Coinbase representative stated. “Deepfakes make convincing scams cheap, fast, and easy to relaunch under a new name the moment one is shut down. This means our defenses must improve at the same rate as the attacks. We are investing heavily in AI across our Risk department to achieve this.”

A Gartner survey of 297 senior cybersecurity leaders, conducted between March and May 2026 and published in September, revealed that 41% of CISOs had encountered at least one social engineering incident involving an employee audio deepfake in the preceding 12 months. Video deepfakes were also prevalent, with over one in three CISOs, or 36%, reporting their organizations had experienced such incidents.

In June, Japanese crypto advocate Mai Fujimoto was deceived by a deepfake impersonating an acquaintance during a Zoom call, which ultimately led to malware stealing access to her accounts. Binance founder Changpeng Zhao (CZ) later cautioned that even video calls can no longer guarantee a person’s identity.

Coinbase AI Foils Musk Crypto Scam, Blocks Millions in Stolen Funds · MissedBlock