Security systems based on artificial intelligence learn from historical activities, incidents, breakthroughs, to build their own models autonomously, without continuous human control.
As organizations continue to grow, the workforce becomes global, diverse and distributed, and as organizations deploy new clouds, on-premises systems, and smart devices, the old static policy model based on a fixed set of contexts (e.g. access control, time, geolocation, device operating system, etc.) is no longer adequate. The number of policy measures is increasing, the context does not take into account the history of users and it will be difficult to protect oneself against future attack vectors.
This is where AI-based security really begins to show its value. Such safety systems draw on historical activities, incidents and injuries to develop their own models autonomously and without constant human control. They are wise when it comes to making their own decisions, and they are astute when it comes to looking at the data, both broad and deep. They are constantly learning and developing with the help of new data so that they are easy to maintain and have an active character. This area has evolved significantly in recent years and is essential for detecting and preventing attacks and injuries. Some of the applications described below.
- The AI /w ML has been used very effectively when huge amounts of data have been sifted through to create identification profiles which are then used to detect not only abnormal behaviour, but also malicious behaviour. This allows administrators to apply adaptive authentication policies, such as B. or only during the privilege/privilege period, to prevent attacks that involve access risks to which a permanent/long-term policy is vulnerable.
- AI is the quality of the data, its completeness and the data science that determines the quality of the analysis (also called modelling). Quality refers to the way the data is cleaned, prepared and shredded for later use. Complexity refers to the different contexts and sources from which the tool collects data. For example, when a user accesses an application, he or she uses a target device (such as a mobile phone) from one location, passes through a firewall network, authenticates the role and then performs certain activities. A good IAM tool is able to collect information from all these contexts (device, location, time, network, directory services, role-based access, etc.). The knowledge generated is then applied to critical resources through adaptive/pro-active guidelines. This approach is important to prevent data leakage.
- From prescribing (providing comprehensive advice) to policy (taking specific measures and automating them) to contain the threats.
- In fact, AI contradicts its basic principles of full autonomy and transforms itself from fragmented, in many cases uncontrolled, learning into a hybrid – a combination of human intelligence and (controlled) contributions with uncontrolled contributions. This leads to a stricter policy, which means fewer false alarms!
- AI is used to orchestrate the configurations of adjacent and affected systems to reduce the spread and impact of irregularities.
- Automated notifications and corrective actions (e.g. to block access or reduce rights). Robotic process automation (RPA) also improves efficiency in this area.
- Use of artificial intelligence for role engineering and identity management. This includes the automated implementation of segregation of duties and risk-based management of business processes.
This position was originally expressed during Quorum Question Time in January 2020. Our architect CPO Lohokare was invited to discuss, among other things, the state of affairs in the field of cyber security, zero trust, artificial technologies, machine studies and security work. Stay tuned as we post more of his comments in our blog!
*** This is a syndicated blog from a network of security bloggers with articles written by architect Lohokare The original post can be found at https://www.idaptive.com/blog/artificial-intelligence-help-avoiding-data-breaches/.data breach artificial intelligence,data security concerns with artificial intelligence,artificial intelligence techniques for cyber security,artificial intelligence vs cyber security,applying artificial intelligence techniques to prevent cyber assaults,how artificial intelligence is changing cybersecurity landscape and preventing cyber attacks,artificial intelligence in cyber warfare,artificial intelligence cyber attacks