Specialists at the Los Alamos National Laboratory (LANL) of the US Department of Energy have developed an artificial intelligence-based system to detect illegal use of supercomputers for mining cryptocurrencies.
“Building on recent computer hacks in Europe and elsewhere, this type of software controller will soon become critical to preventing cryptominers from hacking high-performance computing and resource theft,”Gopinath Chennupati.
The solution developed in the laboratory uses graphs to detect hidden miners. The latter represent all programs in the form of nodes and links between them.
Artificial intelligence compares software graphs, like human fingerprints, and identifies programs that should not work on a given computer.
LANL argues that the use of graphs by the system does not allow it to be deceived by the usual methods that attackers use to disguise miner programs.
According to the laboratory’s researchers, tests using “malicious code for mining bitcoin” showed that the system based on artificial intelligence detected illegal mining operations “faster and more reliable” than conventional analytical tools.
They admitted that the solution does not give a 100% guarantee, but “significantly expands the set of effective approaches” in the fight against cyber intruders.