Using Machine Learning to Make Faster, Smarter Decisions About Insider Threats

Chances are you already have an established process for identifying attackers and blocking external threats. You’ve taken steps to reduce the likelihood of an attack by exercising good cyber hygiene and following key identity and access management (IAM) best practices, such as adhering to the principle of least privilege.

But what about threats from within your organization? An insider threat is when an insider’s credentials and access are used, either deliberately by malicious actors or indirectly by criminals with stolen or acquired credentials, to illicitly obtain sensitive data from an organization. According to a recent SANS survey, 76 percent of security professionals ranked malicious, accidental or negligent insiders as the most damaging threat vector they face.

A Comprehensive Approach to Detecting Insider Threats

Insiders all look the same, making them increasingly difficult to detect. Security professionals need a way to analyze user activity and make intelligent decisions on a case-by-case basis. Traditional approaches to remediating suspected insider threats typically rely on technological solutions, such as blocking a user’s firewall access. This may or not be sufficient, since the user may still have access to resources beyond the firewall. If an insider is suspected of suspicious activity, why not employ tools that help you determine exactly which accounts, data stores and other enterprise resources a particular user can access? IAM solutions provide this information and enable security professionals to block threats fast with a more comprehensive approach.

Before you can make smarter decisions about user access, you’ll need a user behavior analytics (UBA) solution to monitor user activity and assign risk scores to individual profiles. A UBA tool works by examining patterns of user activity while employing machine learning to detect anomalous behavior that could indicate a potential threat. This technology can detect and alert you to behavior that violates security policies, such as attempting to log in multiple times or doing so from an unfamiliar location or at a suspicious time. Machine learning can identify sophisticated breaches in which attackers dwell in the network and gradually move laterally to escalate privileges over time. You’ll know exactly which user is in question and what offenses were committed. This enables you to gain more insight into the incident, automatically suspend the account and revoke access through IAM.

Once you’ve narrowed in on a user exhibiting suspicious behavior, how can you stop him or her? This is where IAM comes into play. Once UBA detects behavior that is out of policy, an identity governance and intelligence (IGI) solution automatically suspends that account. This immediately stops the user during the investigation, blocks him or her from accessing any more information and cuts down on dwell time. That part is critical, since longer the dwell time, the greater the potential for extensive damage. Automation saves crucial time during the investigation and eliminates the need for manual user-centric threat mitigation processes. It can even communicate directly with the impacted user to ask authentication questions via a mobile app. This approach is more comprehensive and effective than blocking firewall access.

Every Second Counts

By using UBA in conjunction with IGI to centrally and securely manage user identities, you can visualize insider threats and automatically control what resources each user can access based on previously established security policies. Depending on the risk score, you can use IGI on a per-resource basis to suspend or shut down access completely.

When responding to insider threats, every second counts. By implementing IAM tools that can automatically respond to suspicious activity, organizations can help minimize damage while security analysts conduct their investigations. The combination of UBA and IGI enables security teams to manage user access with an approach that works silently in the background, only taking action when necessary to minimize disruptions to legitimate business activities.

To learn more, watch this demo video of QRadar UBA + Resilient + IGI integration:

Share this Article:
Michael Bunyard

IAM Product Marketing Lead

Michael leads IBM IAM’s product marketing efforts. He brings over two decades of product management and marketing expertise to this role. Prior to IBM, Michael led marketing for a start-up company in the endpoint detection and response market, product management for Fiserv’s fraud detection solutions, and product marketing for BMC software’s enterprise application integration solutions.