One of the first lessons anyone learns in the Crypto world, apart from “what a private key is” and “what Decentralization is”, is undoubtedly the lesson of “YOU CANNOT TRUST ANYONE”. The issue of security in Crypto has not been resolved, not even at a moderately successful level. There are a thousand ways to be scammed. It is likely that the word “SCAM” is one of the —let’s say— three most used words in Crypto.
People get used to being scared, feeling fear and not being able to trust all too often. One cannot believe content of YouTube videos that are watched, no matter how charismatic and apparently intelligent the YouTuber seems. One cannot trust groups in social networks nor can one believe in every campaign launched by new projects offering the sky and the stars.
Academics from the Imperial College in London have invested months of work in a longitudinal research endeavor that investigates the techniques used to tailor a SCAM and have used the knowledge to design and implement a system of Artificial Intelligence (AI) that can estimate the probability that Internet content corresponds to the patterns of a SCAM.
To this day, AI is known as an attempt to emulate human cognitive processes that allow us to act “intelligently”, and to make decisions that bring us benefit and keep us from danger. Within the many cognitive processes that a biological brain performs are processes of pattern recognition (PR) which can also be programmed using AI.
PR systems are complex programs that can detect patterns in the shape of things, the behavior of an object or a living being, and patterns in language or in communication. At the Imperial College of London, the technologies of PR have been used to predict future emergence of concrete cases of a “Pump and Dump”, through the technology of pattern recognition.
Schemes of a PUMP AND DUMP can be detected by this new PR system, particularly when a group of people get involved in a malicious agreement to “inflate” the price of a cryptocurrency. Together, these people manage to raise the price in an unnatural and inorganic way. Many victims unfortunately fall into the trap because they believe that the currency is experimenting a natural rise in price.
Logically, the insane price inflation happens only for a few hours or a couple of days. When it eventually stops, the price falls precipitously and the victims soon realize the deception; the price was not rising due to any legit reason and the victims ended up paying a highly pumped price for an asset that, in reality, may not have been worth anything. It is well known that scammers usually buy back when the price comes back down.
The AI system being created in London, according to the official sources of the institution, has made it possible to detect and understand many schemes used by scammers. One of the most notorious findings is that AI says that the social platform “Telegram” is the most suitable for this type of virtual cheating. With this AI system, more than 300 suspicious Telegram channels have been analyzed and most show signs of manipulation.
“The researchers found that around 100 Telegram pump-and-dump groups organise two pumps a day on average, ultimately encouraging investors to spend seven million USD per month … They then looked for signs to identify which coin, if any, was about to be pumped by analysing coin features like ratings and market movements. Signs included coin market cap, and unusual fluctuations in price and volume, prior to the pump … Based on these signs, the researchers developed an algorithm that predicts with good accuracy how likely a specific coin is to be pumped before it happens.”
This new PR system is one of the first tools against Crypto mafias that has been tested and is at this moment ready. Programmers around the world have talked about many other types of algorithms, such as those that detect suspicious activity in some blockchain accounts, and those that protect users from the possibility of having private keys stolen. All these innovations represent a hope for users and investors of Crypto in a world full of dangers, hacks and even damages to the integrity of the people.
Do you think AI systems can be programmed to detect scams and help save people from being victims?