← Back to Articles
Focus: Enterprise Security Architecture | Angle: primer

A primer on modern enterprise GRC

A quick overview of modern enterprise GRC.

If you’re doing second or third line evidence gathering with screenshots, you’re doing it wrong. If you don’t know what I mean by second line (of defense) or third line (of defense), you’re more than a decade behind for IT, IT security, GRC, and audit. tl;dr Automate everything.

Modern enterprise risk management uses the Three Lines Model (formerly “The Three Lines of Defense in Effective Risk Management and Control” commonly known as the “Three Lines of Defense”, released in 2013). The second and third line are about controls monitoring and audit. And for each, gathering evidence of controls and their effectiveness is a key requirement.

From the first line, the operations side or the analyst side, you may be used to screenshots, reports, and similar forms of evidence. Need to prove that you use password complexity on your Active Directory controllers? As you’re working down the evidence list, you grab a screenshot, save it with the request number, and move on.

But this doesn’t scale. You’ve got a bunch of pixels that a human needs to review, inspect, know what they’re looking at, and then attribute a pass/fail/partial score. Each time a new image is generated for review, someone in GRC and/or audit has to review the image each time to find the option’s status. And worse than that, it doesn’t provide good resolution—you can only generate & review it so frequently or on a small number of systems before everyone gets tired of such menial work.

Ridiculous.

  • Doesn’t (Get-ADDefaultDomainPasswordPolicy).ComplexityEnabled provide the same information? And it’s scriptable. And it’s directly testable. And it takes less than 15 minutes to find, test, and add to a script.

  • But a true or false by itself isn’t enough to tell someone an answer. We need to know what question was being asked. So outputting the command as well as its response handles that.

  • Concerns about tampering with evidence? Hash the script output (including the command), and store the hash in a separate file, indexed on the script output. Maybe use an HMAC or some other signing function to validate integrity if you need strong assurance.

  • Finally, this approach is evergreen! Run your script quarterly, monthly, weekly, or daily. The computer doesn’t care and it doesn’t get tired of menial work.

For those who work the second or third line, you can process the output of these scripts and validate the answers with your own scripts. There will be both less need and more need for tracking requests, but you can always find the response you need because the command is included.

Finally, a good enterprise security architect sits in a place that straddles the first and second lines. But more importantly than that, they should be ensuring the capability exists to enable this for all three lines. All systems should provide output or APIs that are scriptable or queryable for responding to second and third line evidence requests.

Manual work is for suckers and very old systems.

Bonus points if you create a policy as code system that natively captures and tests each resource for compliance on an ongoing basis for security purposes, then extend access to control owners, GRC, and audit personnel. A single system to pull data automatically and reduce the burden on first line operations personnel saves money while instrumenting a modern enterprise GRC system usable by the second and third lines.