Bots Vs Scambots — How AI can tackle online fraud
By Jonas Sevel Karlberg, CEO at AmaZix
Fraud and scamming have been part of the internet ever since its inception. You know the tricks — you get random emails about Nigerian princes offering millions, or kindly lawyers offering to facilitate the collection of a deceased 3rd-cousin-6-times-removed’s unclaimed multi-million-dollar estate, or some other money-making plan that seems too good to be true (and it always is).
As technology has developed, so too have the fraudsters. When crypto exploded into mainstream public consciousness, it attracted a new breed of white-collar scammers. As these new criminals discovered, short cons carried out using crypto are hard to detect and almost impossible to trace, especially in smaller amounts. In many ways, it’s the perfect crime, and the criminals are all over it with an estimated $670 million lost in crypto fraud in Q1 of 2018 alone.
We often think of crypto scammers as talented hackers, shrouded in mystery, wearing Guy Fawkes masks, and possessing a skillset shared by few others. In reality, crypto cons (like most cons) are actually very easy to carry out, and this is a gift for cyber criminals. As crypto has become more popular, new investors are often unaware of the dangers.
One common crypto scam is an evolution of ‘traditional’ phishing. A fake team member of a crypto project will direct a victim to a cloned website of social media account, which entices them to send their crypto to a wallet address in return for a reward. In most cases, the victim doesn’t realise for days, assuming that their payment has gone to a legitimate project. Only when an acknowledgement email doesn’t arrive, or when the payment doesn’t show up on a dashboard, might a victim notice that they’ve been scammed. By that time, though, it’s too late — the funds are gone, and you have no one to complain to. It’s a simple, easy con that is being used to great success.
But we’re fighting back. As scammers are evolving, we are too, and we’re producing ever more sophisticated tools to fight against con bots using bots of our own. At AmaZix, we have bots working across multiple projects and channels, which delete content and ban users before anyone even notices — once an account is banned in one channel, it’s removed from all our channels.
Bots are catching bots now, and our crypto community moderators are acting as generals presiding over enormous AI battles, with friendly bots designed to keep their users and communities safe by fending off scambots looking to plunder their hard-earned crypto. This AI technology is developing in both power and complexity, allowing security firms to create programs that can trawl huge digital spaces, processing thousands of pieces of information per second. The more information they receive, protector bots teach themselves how to be more effective. As scambots have nowhere near the resources to match this, protector bots have a crucial edge.
As crypto thrives and blockchain is greeted with wider mainstream acceptance, the AI security systems which utilise it will grow in power, offering a hopeful future for crypto communities and the bots that can protect them.
