Microsoft Simulator to help improve AI defense against cyberattacks

Microsoft Authorized Partner



Analyzing the Data

Researchers and safety specialists may use the simulation to examine how cyber attacks operate their way through a network. Microsoft’s blog article about the simulator and AI defense against cyberattacks is a specialized analysis aimed mostly at safety researchers and specialists.

The site article contains quite a lot of jargon which you’ll need to emphasize, however, Microsoft utilizes a beneficial video game event to describe the installation and practice your defense against cyberattacks.

The participant of the match is the broker, the controls it requires are the action, and the greatest reward is winning this match. The very ideal reinforcement learning algorithms may learn successful approaches through repeated expertise by slowly learning exactly what actions to take in each condition of the surroundings. The longer the agents play the sport, the more intelligent they get it.


Training the Software

To put it differently, folks may make a version of computer nodes then have a simulated attacker exploit vulnerabilities and operate its way through this community. This version can subsequently be utilized to develop flea brokers. This installation uses reinforcement learning how to train autonomous agents which could work better than individuals in the games.

This method of security offers valuable insight, however, Microsoft did not use it for simulating traffic. Rather, the plan concentrates on understanding how brokers operate through a community. Microsoft describes towards the conclusion of its website article:

The simulator at CyberBattleSim is also simplistic, that includes benefits: its highly subjective character prohibits direct program into real-world programs, so providing a shield against possible nefarious use of automatic agents trained with that. Additionally, it makes it possible for us to concentrate on particular facets of safety we want to examine and immediately experiment with current machine learning and AI algorithms.

About Us

With this set up, Microsoft expects that individuals are able to research how AI can shield against attacks.

While we (JNS) do not necessarily work with this simulator our team uses our lab to test their training before putting it live on a customer network. Our staff is required to take several training throughout the year and so they are ready to provide any cybersecurity consulting or defense against cybersttacks.


Joint Network Systems
1100 Brickell Bay Drive
Miami Florida 33231
Tel: 866-JNS-NETS