AI-Driven Crypto Crime Has Just Begun – Report by Elliptic – TradingView News

The rise of artificial intelligence (AI)-based cryptocrime is ushering in a new era of cyber threats. A report from Elliptic reveals how sophisticated technologies are being abused for deepfake scams, state-sponsored attacks, and other sophisticated illegal activities.

The report cites a disturbing advertisement for an “unethical” Generative Pre-Trained Transformer (GPT) on the dark web, which claims: “AI has two faces, just like humans.”

This duality is even more evident in the WormGPT advertisement highlighted in the report:

“Experience the dark symphony of code, where there are no more rules and your imagination is the only limit. Together we navigate the shadows of cyberspace and are ready to conquer new frontiers. What's your next step?”

Related: Over 35 fake Elon Musks live streamed the SpaceX launch

Deep insight into deepfake

Elliptic announced that deepfake videos featuring Elon Musk and former Singapore Prime Minister Lee Hsien Loong are being used to promote fraudulent investment schemes.

“Fake videos – or 'deepfakes' – of well-known individuals promoting investment scams targeted the likenesses of Elon Musk, former Singapore Prime Minister Lee Hsien Loong, and Taiwan's 7th and 8th presidents, Tsai Ing-wen and Lai Ching-te.”


The report also highlighted the increasing prevalence of deceptive deepfake tactics used by scammers in social media communities to trick unsuspecting victims into handing over their money.

“Crypto giveaway and doubling scams are increasingly using deepfake videos of crypto CEOs and celebrities to trick victims into sending money to fraudulent crypto addresses.”

Related: US Treasury Department: AI brings enormous opportunities, but also risks to financial stability

USA warns of Korean crime

According to the report, Anne Neuberger, US deputy national security adviser for cybersecurity and emerging technologies, also addressed growing concerns about AI crime.

Neuberger warned that AI is also being misused for purposes other than everyday fraud:

“Some North Korean and other state and criminal actors have been observed attempting to use AI models to accelerate malware creation and identify vulnerable systems.”

Related: Meta faces criticism in the EU for using AI data without users' consent

AI crime is not worth it

The report details the extent of artificial intelligence abuse seen in darknet forums.

“Elliptic has identified chats on numerous darknet cybercrime forums discussing the use of LLMs to reverse engineer seed phrases for crypto wallets, bypass authentication for services such as OnlyFans, and offer alternatives to image manipulation services such as DeepNude that allow for image stripping.”

As activity on the darknet increases, so does the risk of a crash, as demonstrated by the recent arrest of the owner of a darknet marketplace on May 18 in New York.

The 23-year-old was charged with owning, operating and profiting from a $100 million darknet drug marketplace after the FBI tracked his crypto transfers.