A scam involving fake YouTube tutorials for creating AI trading bots deceived 224 people, who lost a total of 274.6 ETH (around $517,000). Instead of using traditional phishing links or wallet permission requests, the attackers convinced victims to deploy malicious smart contracts themselves, which then stole their funds.
According to a TRM Labs report published on September 14, the operation disguised malicious Ethereum contracts as automated trading tools created with Anthropic Claude. This allowed the scammers to steal funds without relying on conventional phishing methods. The blockchain intelligence firm tracked 234 contracts deployed by victims, although the campaign affected 224 people because some participants created more than one contract.
Key statistics: 274.6 ETH (~$517,000) stolen; 224 victims deployed 234 malicious contracts; median loss of 1 ETH per incident; funds transferred to 6 collection addresses; 9 similar YouTube videos featuring AI-generated presenters.
Instead of directing users to a page that immediately requested access to their wallets, the operators presented the scheme as an educational process. Victims found the videos, followed the instructions and performed every on-chain step themselves. TRM identified nine nearly identical YouTube tutorials presented under different creator identities.
AI-generated virtual presenters and voiceovers made the videos appear to be independent guides, while each tutorial promised to help viewers create a fully automated crypto arbitrage bot using Claude. During the videos, users were instructed to copy code and open a compiler website selected by the presenter. Some of the sites copied the design of Remix — a widely used browser-based development environment for writing and deploying Ethereum smart contracts.
The victims then connected their wallets, compiled what appeared to be trading software and deployed the resulting contracts. Because users initiated and approved every action themselves, the transactions looked different from attacks in which a fraudulent website requests direct token permissions or an unclear signature. Funding the newly deployed contracts completed the trap.
Users believed they were providing capital that the bot would use to exploit price differences between trading venues, but TRM found no arbitrage system or AI functionality in the malicious contract variant it examined. In effect, the victims became the deployers of the malicious code themselves, making the entire operation appear legitimate from the blockchain's perspective.
In one version of the scheme, a backend script ignored the source code that victims pasted into the compiler. Instead, the website retrieved a separate contract from a server controlled by the scammers and prepared the replacement for deployment. As a result, the clean code displayed in the browser was never deployed on-chain. Victims saw one program on their screens while their wallets deployed another, preventing them from verifying the actual contract simply by visually inspecting the compiler window.
The replacement contract could accept ETH deposits, matching the expected behavior of a trading bot that needed funds to operate. However, once its balance exceeded 0.05 ETH, the contract was configured to transfer the money to an address controlled by the operators when the user selected either the Start or Withdraw function. Both buttons therefore served the same purpose despite their labels resembling normal bot controls.
Pressing Start did not activate a trading strategy, and pressing Withdraw did not return the deposited funds to the user. According to TRM's findings, no AI model interacted with the deployed contract. Claude branding was part of the marketing pitch, while the on-chain code merely accepted deposits and transferred qualifying balances to the scammers.
Technical deception: A backend script replaced the displayed code with a malicious contract. Victims saw harmless trading code, but their wallets deployed a trap contract. The Start and Withdraw buttons both transferred funds to the scammers when the balance exceeded 0.05 ETH.
This method also reduced the likelihood that standard wallet protections would interrupt the process. A wallet could accurately show that its owner was deploying a contract, sending ETH to it and later calling one of its functions, but it still lacked the context needed to determine that the tutorial and compiler had misrepresented the code. This made the attack particularly deceptive — every action appeared legitimate from the wallet's perspective.
Traditional crypto phishing campaigns often depend on cloned domains, poisoned search results or requests for broad token permissions. Blacklists and wallet simulations can sometimes identify a known malicious address, fraudulent domain or transaction that gives an attacker control over existing assets.
The AI trading bot operation took a different approach because each victim became the deployer of a newly created contract. A fresh address would not necessarily appear on existing blacklists, while the wallet owner authorized the deployment and funding transactions without surrendering a seed phrase. This bypassed traditional protections that rely on known malicious addresses or suspicious permission-request patterns.
In July, crypto.news explained how drainer tools typically abuse legitimate blockchain permissions. Such tools often convince users to approve a malicious contract that can then transfer tokens, while the blockchain processes the action as authorized. The campaign described by TRM moved the deception one step earlier by controlling the code generation and deployment process.
Instead of asking victims to trust an existing contract, the tutorials convinced them that they were creating the software themselves. This psychologically lowered their guard — people believed they controlled the process when, in reality, they were following a carefully designed deception script. AI-generated presenters added another layer of credibility by creating the illusion of professional educational content.
The method also exploited growing interest in AI technology within the crypto community. The promise of creating an "automated arbitrage bot with Claude" sounded technologically advanced and attractive to traders looking for an edge in the market. The scammers used the hype surrounding AI to build a convincing narrative that overcame the usual skepticism toward "easy money."
For U.S. users, the FBI's Internet Crime Complaint Center (IC3) accepts reports related to crypto scams and other cybercrime. The bureau says complaint data can help investigators identify related cases, track emerging methods and, in some situations, freeze stolen funds. The FBI recorded $16.6 billion in reported internet crime losses in 2024, compared with $12.5 billion in 2023.
The agency advises victims to file reports even when they are unsure whether their complaint fits a specific crime category, because the information may be forwarded to federal, state, local or international law enforcement agencies. Although fund recovery is unlikely, reporting helps law enforcement track patterns and potentially identify the operators behind the scheme.
Developers and educational platforms should clearly label AI-generated content and provide code verification before it is used. For users, it is critical to understand that deploying smart contracts is a serious action with irreversible consequences and requires the same level of caution as handling private keys.
The industry also needs better tools for detecting schemes like these at the compiler and development-environment level. Future versions of Remix and other IDEs may include warnings about suspicious code patterns or integration with real-time auditing services. Until then, vigilance and skepticism remain the best defenses against increasingly sophisticated scams exploiting the hype around AI technology.
“In the world of cryptocurrency, if an offer sounds too good to be true, it almost always is a scam. AI technology does not change this fundamental rule — it simply makes the deception more convincing.”
— Andreas Antonopoulos, Bitcoin and security expert
