Coinbase AI defense blocks $4.7 million from being sent to deepfake Elon Musk scam

LongbridgeAII'm LongbridgeAI. Here is the key information I extracted
- Coinbase's automated AI defense system successfully blocked $4.7 million from being sent by customers to a deepfake Elon Musk crypto scam livestream.
- The fraudulent YouTube broadcast tricked hundreds of users into voluntarily authorizing transactions to fake Tesla giveaway addresses during a product demonstration.…
Disclaimer

Coinbase (COIN) said its AI-powered scam defenses stopped $4.7 million in funds that customers tried to send to scammers who used an Elon Musk deepfake to promote a fake Tesla crypto giveaway in a livestream. The company said the system prevented about 91% of the attempted transfers, with no humans in the loop.

According to Coinbase's blog post published on Tuesday, the incident occurred during Tesla's "flying car" demonstration last month. Scammers created a YouTube livestream that resembled Tesla's official broadcast, featuring a deepfake of Elon Musk. It also featured on-screen wallet addresses and a QR code, offering to double any cryptocurrency sent. More than 500,000 people watched the stream, according to Coinbase.

Within roughly two hours, hundreds of Coinbase customers attempted to send millions of dollars to the addresses shown.

"The part that makes this hard: nobody was hacked," Coinbase said. "No Coinbase account was broken into, no password stolen, no credential phished. Hundreds of customers, many of them long-time users, decided on their own to send money to a stranger."

While most platforms have adopted real-time fraud detection systems, Coinbase explained that it is difficult for them to stop transactions that customers were actually tricked into authorizing.

Closing the loophole

Coinbase said its AI defense system closed that loophole by flagging the activity in real time using machine-learning models that score risky sends, including those matching giveaway patterns, brand impersonation, and newly created wallet addresses.

Then, it dispatched a dedicated scam-prevention agent to examine the context of the transaction, explain how the scams are structured, and ask questions that deconstruct the story scammers were pitching.

"What's notable is that the agent had never seen this scam. Deepfake giveaways tied to a live product launch weren't in its training," Coinbase said. "It recognized the shape of the scheme anyway, a promise to multiply funds, the urgency of an event that would end in hours, a transfer that could not be reversed, and it ran from the first flagged send to the last, with no humans in the loop."

Coinbase said it will continue to invest heavily in AI defense to create models that keep pace with new scam types through AI-driven research.

U.S. Business AI Use

Source: The Block The copyright of this article belongs to the original author/organization.

The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.