How can artificial intelligence support penetration testing?

Most processes in for the penetration-testing cycle require detailed knowledge, time and human resources.
While the are sophisticated scripts for the reconnaissance and various exploits, creating a detailed plan of the attack path can be complicated and laborious. The use of an enforcement learning algorithm can help penetration-testing identify the various attack vectors and provide a detailed overview of the system landscape. This can automate important aspects of the process and make it more efficient.

We like show an overview, on how reinforcement learning can be integrated into the penetration testing process to gain automated access to a system landscape.
To achieve this, we show approaches how an AI can be used for lateral movement within the system landscape to subject an entire landscape to the penetration-testing process.