AI to Fight Crypto Scams on X
American tech developers have launched an AI-driven tool adept at pinpointing deceitful accounts touting free token handouts on social platforms. An exhaustive account of its accomplishments has been unveiled.
Researchers from the University of San Diego have rolled out a system, named GiveawayScamHunter. This state-of-the-art tool is designed to sift through mountains of text in mere minutes, cross-referencing it with established language patterns to pinpoint manipulative NLP cues.
Given that these deceptive practices have become rather formulaic, con artists stick to tried-and-true methods:
- promises of quick earnings without effort;
- invitations to join token giveaways;
- stories of fortunate users who've already cashed in;
- warnings that the opportunity will soon disappear.
So, for GiveawayScamHunter, tracking this algorithm in posts isn't an issue.
To underscore the prowess of their invention, the developers showcased its pilot run outcomes. The data was shocking: this digital sleuth identified a whopping 88,000 con artist accounts set up on X from June 2022 to June 2023.
But that's not all. The Artificial Intelligence was able to extract:
- 125 wallet addresses linked to criminal groups;
- 327 free internet domains owned by them;
- 365 confirmed victims who lost money (note, this figure pertains only to Twitter accounts; the real victim count? One can only guess);
- approximately 100,000 lists of fraudulent giveaways.
Based on preliminary data, the stolen funds during the studied timeframe amount to over $870,000. But this is a conservative estimate. The minds behind GiveawayScamHunter believe the true figure might be much higher, especially when you consider lists with more than 100 participants represent almost a quarter of the total.
The report also provides an analysis of which cryptocurrencies are most commonly involved in scam messages on Twitter. Ethereum (ETH) takes the top spot, followed closely by Bitcoin (BTC), then Polkadot (DOT) and Shiba Inu (SHIB). Ada (ADA) is least commonly used in these deceptive offers.
The research revealed that a concerning 13.5% of all crypto giveaway promotions on Elon Musk's platform are fraudulent (accounting for 95,111 out of 705,576). What's more alarming is that almost half of these scams remain active, continuing to deceive users.
This situation surely rings alarm bells for the singular owner of the "X" brand. Elon Musk envisions transforming the social media landscape but struggles to address the rampant scams on his platform.
And university scholars aren't the only ones leveraging AI to root out deception. Scammers are also capitalizing on these technological advances, expanding their outreach campaigns and building impressive follower counts in the tens of thousands. They're innovatively deploying AI-driven chatbots and virtual assistants to interact more effectively with potential targets.
