“Based on recent computer break-ins in Europe and America, this type of watchdog software will soon be vital to prevent cryptocurrency miners from infiltrating high-performance computer facilities and stealing valuable computing resources”, said Gopinath Chennupati, researcher at Los Alamos National Laboratory. “The artificial intelligence (AI) deep learning model has been designed to detect the abusive use of supercomputers specifically for the purposes of cryptocurrency mining.”
Cryptocurrencies, such as Bitcoin, are forms of digital money. The cryptocurrency miners «dig» digitally for coins by performing intensive calculations.
Legitimate “cryptocurrency miners” often amass vast computer clusters to dedicate to collecting digital cash. But others looking to get rich quick are trying to hijack supercomputers, trying to avoid detection, of course. The new artificial intelligence (AI) system is designed to spot these actions in action by comparing graph-based programs , which are like fingerprints for software
All programs can be represented by graphs made up of nodes connected by lines and loops. The new AI system compares the contours in a program with a list of graphs for programs that are allowed to run on a given computer.
However, instead of finding a match to a known criminal program, the system checks to determine if a graph is among those that identify programs that are supposedly running on the system.
The researchers tested their system by comparing a known, benign code with an abusive Bitcoin. They found that system detected the illegal mining much faster and more reliably than conventional non-AI analyses.
Because the approach is based on graph comparisons, it cannot be fooled by common techniques that cryptocurrency miners to obfuscate their codes, such as variable cloaking and comments intended to make their codes look like legitimate programming.
While this graph-based approach may not offer a completely effective solution for all scenarios, it significantly expands the set of effective approaches that cyber detectives must employ in their ongoing efforts to quell cybercriminals.
Based on the recent changes in the so-called “computer break-ins”, such software will soon be vital for preventing the intrusion of cryptocurrency miners into high-performance computer installations and the theft of valuable computing resources.
