Insider data breaches can be a real problem in businesses. One way to address this issue is by implementing an identity and access management solution.
What tips do you have for ensuring that one's identity and access management solution is effective?
The simplest and most common activity for every insider threat action is the logon. Nearly all threat actions require a logon using internal credentials. Endpoint access, lateral movement between endpoints, external access via VPN, remote desktop access, and more all share the common requirement of a logon.
Remember also that almost every external attack eventually looks like an insider. The use of compromised internal credentials is the most common threat action in data breaches.
To ensure the best out of any access management solution, think around five primary functions – all working in concert to maintain a secure environment.
Two Factor Authentication – Regulating user access involves authentication to verify the identity of a user. But authentication using only a strong user name and password doesn’t cut it anymore. Two-factor authentication combines something you know (your password) with something you have (a token or authenticator application).
Access Restrictions – Policies can be added on who can logon when, from where, for how long, how often, and how frequent. It can also limit specific combinations of logon types (such as console- and RDP-based logons).
Access Monitoring – Awareness of every single logon as it occurs serves as the basis for the enforcing policy, alerting, reporting, and more.
Access Alerting – Notifying IT - and users themselves - of inappropriate logon activity and failed attempts helps alert on suspicious events involving credentials.
Access Response – Allows IT to interact with a suspect session, to lock the console, log off the user, or even block them from further logons.
The potential insider threat scenarios that are now thwarted include:
With experience in both IT and Audit, I can say the answer most often leads to a tried and true combination of preventative and detective mechanisms/controls. These two methods though very different help with achieving the goal of minimizing breaches and detecting them so the right action is triggered when a breach does occur. Since every business has to place on a scale cost vs risk, unless the business has endless monies, there will be some risks too expensive to prevent so you must have the means to detect and then react with the goal of minimizing the exposure and learning from it.
A ridiculous example but proves my point: Every employee has a second or third employee watching and validating every action carried out by the first employee to ensure no data breaches. So the risk is minimized and maybe even eliminated but the cost is more than most companies will ever contemplate. I will leave alone the topic of collusion since that is more than we can explain in this short answer. Now remove the 3rd watcher person and reduce the 2nd by 50% to save money but scope the first person's actions. If the first employee's actions are limited by the roles assigned (in a system or manual), the activity carried out by the employee is controlled and scoped which in turn limits risk. The remainder is added to detective mechanisms such as DLP in a system or even a human reviewing (maybe sampling) the first person's activities.
It is a roundabout way to say, you need a combination of both types of controls where access is scoped and monitored. Where the availability of the data is limited to the degree cost-effective and then the less costly but less reliable detective means are used.
The premise of any effective Identity and Access Management solution is that 100% "Trust" exists. Unfortunately, trusting someone to the "keys of the kingdom" is best left to Hollywood, while ensuring the business stays afloat in the real world requires that a robust zero trust mechanism be implemented. New employees, whether experienced or fresh out of school, do not have the luxury of developing the level of trust that can be deemed "100%".
There are easily a dozen low hanging fruit and I would start with the none tech vector: data owners and stewards. Then comes the education and policy dissemination of the company’s stand on data loss. After a move to the tech implementation to detect common signs such as DLP identifying when large and frequent data transfers via email, copy to external drives which include cloud and thumb.
Once you've selected the right solution for your business, you need to make the implementation a formal project and involve all key stakeholders, including those from the business, not just IT folks. Identify all of your information assets, classify them based on sensitivity and criticality (e.g. Public, Internal Use Only, Confidential, and Restricted), then create rules for the granting, revocation and modification of access to those assets. Once that is done and everyone is aware of the policies and procedures governing access, you can implement the solution accordingly. Post-implementation you will want to have a process in place for periodic review of access based on applicable regulatory, audit and security requirements. You may have to create custom reports if the canned reports are not sufficient. Data owners should be involved in the review since they are usually in a better position to determine if individual's access is still legitimate.
Bearing in mind that 100% trust is impossible, it is best to get to zero trust as soon as possible within the confines of your company's risk appetite and with the best tools your company can afford. There are many Identity and Access Management products and services out there - choose wisely and carefully.