You’re the weakest link in fighting crypto fraud, but AI-powered help is here


The proliferation of crypto fraud and scams reminds us that individuals are the ultimate decision-makers when it comes to engaging with criminals. However, new tools may help targeted victims make better decisions and keep their funds safe.

We spoke with Jonathan Anastasia, president of crypto security specialist CUBE3.AI, to discuss current trends, emerging threats, and possible solutions in the fight against fraudsters and scammers in the crypto space. The company leverages AI and machine learning to prevent crypto fraud in real-time.

Risks of the ‘easy buck’

ADVERTISEMENT

Anastasia, a West Point graduate and former US Army officer, stresses the importance of educating and informing crypto users about potential threats so they don’t engage in fraudulent schemes: “The individual is always the weak link, right?”

Regular users, especially newcomers, often lack the technical knowledge or motivation to research fraud prevention, which makes them vulnerable to online scams.

“If they [a website] only ask you for your library card, maybe that's not where you should be buying your crypto. You should want a level of friction to get on board,” Anastasia emphasized, reminding us of the adage: if something is too good to be true, it probably is.

According to Anastasia, bitcoin (BTC) might be a safer bet when it comes to investing in specific crypto assets. As the largest cryptocurrency, BTC benefits from a wealth of protection and knowledge from exchanges.

Meanwhile, as recent events have shown, altcoins carry higher risks, as their infrastructure is less trusted and more prone to scams and rug pulls.

Linas Kmieliauskas Stefanie Gintaras Radauskas Niamh Ancell BW
Don’t miss our latest stories on Google News

Emerging sectors like real-world assets (RWAs) – tokenized traditional asset classes such as bonds, stocks, and real estate – also broaden the playing field for criminals. Stopping the illicit transfer of tokenized assets and recovering them if they fall into the wrong hands remains a challenge.

“When people think they can make an easy buck, they tend to make mistakes,” Anastasia said.

ADVERTISEMENT

However, he also noted the need to balance friction and user-friendliness, ensuring that security measures don’t discourage users from joining platforms.

Compliance is not enough

Companies’ attitudes towards the crypto space have evolved. Many businesses, once indifferent to their customers’ crypto activities, now recognize that ignoring the risks is no longer acceptable.

Anastasia explains that when a bank’s customer sends money to a fraudulent crypto exchange, the victim is likely to turn to the bank for a chargeback. As a result, companies are seeking ways to protect both themselves and their customers by creating a safer environment.

This safer environment goes beyond compliance, which, among other things, requires that companies don’t engage with sanctioned entities. Increasingly, it's about preventing crime before it happens.

CUBE3.AI provides clients, ranging from traditional financial institutions to crypto exchanges and decentralized finance (DeFi) protocols, with a score to assess potential risks. This allows them to stop suspicious transactions in real-time.

“The customer has enough information to say, 'You know what? That's probably a scammy website, or that's a scam address. I'm not going to send them that money,'” Anastasia explained while reminding us that the ultimate decision still lies with the sender.

“But at the end of the day, we also recognize that the customer we work with will know their customer better than us,” Anastasia added, noting that companies may proceed with transactions deemed risky because they have more nuanced information about their clients.

“Sometimes, it starts with a one-on-one conversation on social media or a website. That’s where these things begin.”

Jonathan Anastasia.

Where the fraud begins

ADVERTISEMENT

Victims in the Web3 world, which is sometimes used interchangeably with the crypto industry, are often targeted on Web2 platforms like social media networks and dating apps. According to Anastasia, it's all about human interaction.

“Sometimes, it starts with a one-on-one conversation on social media or a website. That’s where these things begin,” he added.

To address this, the company expanded its services in September by scanning Web2 content, including forums, social media platforms, compromised addresses, phishing sites, and dark web activities. This integration of data from both Web2 and Web3 sources provides a composite risk score, which could help users make more informed decisions.

This approach could also help protect against social engineering schemes, such as the increasingly popular “pig butchering” scams, and other fraudulent activities.

AI vs AI

AI-powered fraud is A growing threat, making it harder to verify whether one is engaging with the right person. Anastasia explained that while risky interactions can sometimes be identified through blockchain and Web2 data, it's becoming more difficult when AI is used to fake a voice, for example, pretending to be a bank representative or relative.

This is why solutions that address verification, attestation, and authentication problems are likely to be in high demand.

CUBE3.AI has been exploring the Ethereum Attestation Service (EAS), an open-source infrastructure for making attestations on-chain or off-chain. According to Anastasia, using blockchain or services like EAS can create a source of verification that a document came from the right place, though it might not be an absolute source of truth.

Meanwhile, Anastasia noted that addressing these issues by giving Web2 platforms full access to personal data and interactions could raise privacy concerns. He suggests that a unified front against AI-powered threats might be a more successful solution.

According to Anastasia, AI-powered fraud is currently ahead of AI-powered protection. However, in this ongoing cat-and-mouse game between fraudsters and defenders, AI-powered protection might eventually overtake AI-driven threats before criminals find new ways to bypass them.

ADVERTISEMENT