When you read about a data breach in the news, it’s never a case where the 10,000 records stolen went missing. Instead it’s usually brought to light via some external source (like a number of customers complaining their credit cards data was misused just after a purchase on your website). Once an organization comes to grips with the fact they may have a problem, it becomes time to put the incident plan into action.
The incident response team knows there likely has been a compromise, but, again, it’s not like there’s a giant hole in the side of a server or something to identify exactly when, where, what, and how the breach occurred.
So, what exactly does a compromise look like?
It’s not a simple question to answer, as identifying a compromise takes a material amount of detective work to review just about everything that has occurred on your network, working backwards from the breached data in question.
Compromises show themselves as irregularities in user activity in a number of forms:
- Admittance – External threat actors need to somehow get into the network. The use of phishing attacks and malware continues to be a predominant tactic.
- Access – Threat actors need access to your network, to your systems and, ultimately, your data. Misuse of credentials to logon and access systems, applications, and data are so common a threat tactic, that these are some of the most likely points at which to spot compromise.
- Applications – The applications your users normally use are far different from the ones an external threat actor uses. Threat actors leverage scripting, PowerShell, downloaded hacking tools, and administrative tools to gain control, establish persistence, and maintain stealth on your network.
- Actions – Insider threat actors will use the same applications they normally use to do their job; it’s only their purpose and intent that have changed. So, insider compromise looks more like irregular timing and use of the network itself (e.g. logging on a 11pm on a Saturday night), application functionality (e.g. excessive accessing of sensitive data), and data transfers (e.g. printing, emailing attachments via webmail, etc.)
- Amounts – Usually the goal of an attack – whether internal or external – is to exfiltrate data. So, irregularities in the amount of data being accessed, copied, transferred to USB, uploaded to cloud storage, etc. – in comparison to what’s normal for a given user – may indicate a compromise.
- Alignment – Because no threat actor wants to be found out, an underlying goal is to attempt to perform their dastardly deeds under the guise of normal everyday network and user activity. Sneaky methods like running file copy traffic over HTTP (which is allowed through your firewall) demonstrate a misalignment of network traffic to the protocol used. Since this almost never happens (with the common allowed exceptions like running a remote web meeting over HTTP), this is a great example of what a compromise may look like.
Identifying a Compromise
There are two ways in which the activity irregularities above are used. Either they are the focus of an investigation – potentially months or even years after an actual breach – or are leveraged via proactive monitoring and auditing to spot compromises the moment they happen. If your organization falls into the former and not the latter, you’ll soon find it’s an arduous – and, something, impossible – task of trying to audit systems, applications, and data access. In many cases, an audit trail simply doesn’t exist.
So, it becomes evident that putting proactive measures in place to watch for compromise empowers IT to both thwart threat actions early on in the process, as well as create the needed audit trail should a successful breach occur.