Security measures can often be placed into one of two categories: defensive and proactive. The first is concerned with defending systems against malicious attacks, usually after they occur. The second is more concerned with proactively managing to block and deflect attacks before the become an issue.
Now, however, a new AI system is actually taking a different stance altogether: taking the fight back to email scammers themselves.
Email scams usually work in a very similar way. Attackers will send out an email in attempt to gain information through social engineering or they will send out an email with malicious links in an effort to infect the target system with malware, spyware, and other similar goodies.
Attempts at social engineering are very rarely well thought-out, of course. In many cases, the person on the other end of the line will not even speak the language very well, whether that is English or whatever other language they speak in the country the attacker focuses on at the time.
One thing they all have in common is persistence and insistence. If a victim is unaware of the scam and actually replies to the attacker, the would-be scammer can sit there for hours attempting to convince the victim of their legitimacy and why they should be transferring their money to a Nigerian prince.
Netsafe’s Re:scam capitalizes on that in the most ingenious way: it can actually converse with the scammers to waste their “precious” time.
The way it works is pretty simple. Users can forward scam emails to the AI and, with the aid of machine learning, Re:scam can generate conversations with them on the spot.
The beauty of all this, of course, is that scammers waste their time on an AI system while they could be targeting actual victims instead. The current record for the scammer-fighting AI 1000 simultaneous conversations, which means that 1000 scammers were denied a victim at the same time.
As is usually the case with AI and machine learning, the system will also become better as time goes on, allowing it to actually trick scammers even more effectively. The system already has multiple variations to make itself as believable as a real person, and it will only become better with time and precious data.
To check out a demo of the AI in action, make sure to watch the video down below. If you get Skynet vibes from it, fret not; this AI is definitely on the side of good.