Identity and access management dynamic control and remediation

A system for identity and access management, comprising one or more processors and one or more non-transitory computer-readable memories coupled to the one or more processors and configured with instructions executable by the one or more processors to cause the system to perform operations comprising:

• obtaining historical user behavior data corresponding to a plurality of user types;
• generating, using a machine learning model, a plurality of user-type specific baselines for the plurality of user types based on the historical user behavior data, wherein each user-type specific baseline comprises a user access pattern of a specific user type;
• receiving an access request from a current user;
• determining a user type of the current user based on a static rule;
• determining an account specific baseline based on the user type of the current user and the plurality of user-type specific baselines;
• obtaining current user behavior data associated with the current user;
• comparing the current user behavior data to the account specific baseline;
• granting, if the current user behavior data satisfies the account specific baseline, an access to the current user;
• denying, if the current user behavior data does not satisfy the account specific baseline, the access to the current user;
• inputting, in real-time, the current user behavior data associated with the current user into the machine learning model;
• and
• updating, in real-time, one of the plurality of user-type specific baselines corresponding to the user type of the current user using the machine learning model based on the current user behavior data associated with the current user.

A non-transitory computer-readable storage medium configured with instructions executable by one or more processors to cause the one or more processors to perform operations comprising:

• obtaining historical user behavior data corresponding to a plurality of user types;
• generating, using a machine learning model, a plurality of user-type specific baselines for the plurality of user types based on the historical user behavior data, wherein each user-type specific baseline comprises a user access pattern of a specific user type;
• receiving an access request from a current user;
• determining a user type of the current user based on a static rule;
• determining an account specific baseline based on the user type of the current user and the plurality of user-type specific baselines;
• obtaining current user behavior data associated with the current user;
• comparing the current user behavior data to the account specific baseline;
• granting, if the current user behavior data satisfies the account specific baseline, an access to the current user;
• denying, if the current user behavior data does not satisfy the account specific baseline, the access to the current user;
• inputting, in real-time, the current user behavior data associated with the current user into the machine learning model;
• and
• updating, in real-time, one of the plurality of user-type specific baselines corresponding to the user type of the current user using the machine learning model based on the current user behavior data associated with the current user.