Offensive security and AI are powerful tools.

The purpose of AI in network security is to support the people and uncover if somebody is telling wrong information. The mission of those things is to see if something works the wrong way. 

The most important thing in data security is to know what kind of attacker might attempt to steal data. In all companies are things that somebody wants. And one of the things is money. The "bad guys" might just want to get the money from the bank account. But in some firms, some secrets are touching the security of the state. 

But all development and research are interest in industrial spies. Those people want to use other people's work without paying licenses and other payments. They want to steal the work of other people and make money with that. 

So knowing the data that is wanted to secure makes sure that the data security team can predict what kind of attackers they are facing. If the attacker is governmental like some intelligence agency, making crime reports does not affect. Things like mobile solutions are interesting the governmental operators. So the company must not necessarily develop missile systems or stealth aircraft for facing the attack, that comes from somewhere far away from the east. 

The reason why things like data security are automatized is that it should eliminate the humanitarian effect. But the fact is that these kinds of systems are always hybrid tools. Artificial intelligence is working as an assistant for people. Who is responsible for the security of the systems. Artificial intelligence can collect data about things, is somebody making something wrong. 

If some people are making the inspection tour around the house. That system sees are they visit in all places that are marked in the list. If those people are not following orders, they are at risk for security. The offensive security means that the workspaces are searched for eavesdropping. 

AI-based network security is the new and powerful tool for securing data. The AI can see things what the normal security manager doesn't see or even realize. There are many layers that AI can manage very effectively, and it can make the profile of all workers by collecting data about things, what people are doing. And if the data, that the user search is changed that thing might cause that something is wrong. 

The weakest point in all data security is humans. Even the most secured systems are vulnerable if the humans who use them are somehow leaking the passwords. In the most traditional model, the "bad guys" are just hitting the person, and force their victim to tell the password and username. That's why face-to-face meetings are also important in working life. And of course, security personnel might follow the workers to home and observe their meetings if there is something wrong. 

During those meetings, the person can tell that if there is something wrong and tell about the strange people who may hang around some workers. Also, there is good to see what kind of questions people are asking. Is some person suddenly start to show interest in the hobbies of the family members of the key persons of the company? Or is some person's behavior changed somehow?

During the network meetings, the AI can see is there a possibility that some people who are involving the meetings are virtual. If the voice of the participant seems somehow synthetic, the system can send the warning to system supervisors. Even the most intelligent systems are requiring people, who are keeping it clean from bugs. The modern eavesdrop systems are not listening only to people. They are also listening to the internet. There is the possibility to put the small WLAN support station to the network connection cable. And route all data through that station. 

When we are thinking about the changing of hacking. Data that is collected from the network must continue all the time. The network security people must always see, what are the settings of the routers. Is some router programmed to double data, and of course the connections must be check because there is the possibility that somebody has installed the "bandit station" to network. 

But new technology brings new threats. The quadcopters can fly to the next roof, and they can equip with cameras, which can see screens that are seen from the window. They can also use laser microphones for uncovering secrets. Laser microphones are seeing the oscillation of the windows and that oscillation transmits speech to the surveillance operators. 

Modern eavesdropping systems cannot detect so easily as traditional systems. They can record data in passive mode, and only when they get radio commands, those systems are releasing the data. So that means they are hard to detect by using traditional methods.