Instead of focusing on breaching the perimeter, attackers have increasingly shifted to compromising the human layer.
With the fallout of the data breach at the Office of Personnel Management still in the news cycle, now is a good time for federal organizations to reflect on the state of their own security and the sophistication of their enemies.
There are many security analysts out there who are more than willing to give their two cents on what the OPM did wrong, but we can all agree that the department was woefully ill prepared to address the tactics of their adversary.
The reality is that most attackers are not breaking into networks; they are just logging in. Defenders are waiting for threat actors to hack through the firewall, but it is easier and more effective for attackers to compromise the credentials and access privileges of organization insiders, then operate with all of the privileges of legitimate users. They are turning innocent users into insider threats.
OPM, which handles security clearances and houses sensitive information on millions of current, former and potential government employees, is no exception to this trend. The attackers gained access to the OPM network by compromising and stealing credentials from KeyPoint Government Solutions, a contractor used to conduct background checks. This gave the attackers the insider privileges needed to persist on the OPM network and steal critical information such as Social Security numbers and medical records of federal employees.
The moral of the story is that traditional perimeter security is no longer adequate. Organizations of all kinds need to be prepared to address insider threats, whether it is a malicious employee or their compromised credentials. The attackers will get past the gate eventually. The only way to stop them then is to monitor internal network activity – leaving the threats with no place to hide – and respond to security events before they become breaches.
OPM became a victim because they had minimal internal security. A report from the Inspector General in November 2014 outlined some of the security shortcomings of the office:
- There was no comprehensive and maintained list of servers, databases and network devices.
- Not all OPM systems were adequately monitored.
- Multi-factor authentication was not required to access OPM systems.
Furthermore, the critical data, including Social Security numbers, were not encrypted because it was not feasable on systems as old as OPM's, according to agency CIO Donna Seymour.
Today's advanced threats have adapted to traditional security measures by focusing on compromising legitimate access credentials. If OPM had been prepared to deal with a threat from inside the network, this breach may have been prevented.
What is behind this shift in tactics?
Over the past few decades, organizations have been pumping billions of dollars into strengthening their perimeters and managing vulnerabilities. Meanwhile, the rise of remote access and personal mobile devices have broadened the threat surface and brought more sensitive data in contact with the internet.
Instead of focusing on breaching the perimeter, attackers have just shifted to compromising the human layer, which is more reachable now than ever before. In many organizations, employees have generous access privileges and the ability to log into the network remotely, which means attackers have ample opportunities to utilize compromised credentials. Additionally, personal information about employees is widely accessible via social media sites like Facebook or LinkedIn, which gives attackers better insight into how to fool them.
Here's a hypothetical scenario: An attacker has managed to track down an employee named Mark on social media. Mark likes to talk about his job and his favorite online poker site. The attacker sends Mark an email posing as a representative from the poker site with an attached brochure on new services, complete with malware. Mark opens the attachment without a second thought, and a few days later the malware sends keystroke information – including his VPN login credentials – back to the attacker.
Now Mark has effectively become an insider threat. Unfortunately, no matter how strong our castle walls are, users who appear legitimate are able to walk right through the front gate.
How to catch an insider threat
Since it is nearly impossible to stop an insider threat at the gate, early detection is key. Fortunately for us, an attack requires more than the initial breach. The perpetrators still must execute a number of steps before their goal is complete, and we can stop them at any point in this process.
The first thing an organization needs to catch an insider threat is network visibility. If firewalls are armed guards at the gate, visibility is the security camera monitoring the inside of the building. Telemetry should be collected from every area of the network. Internal traffic data, access logs, policy violations and more must be watched continuously for suspicious activity. Leaving an area unmonitored gives your attacker a place to hide on your network.
Know what a regular day looks like on your network. Know how much traffic to expect, who is expected to access sensitive information and what applications are used in the day-to-day business operation. Anything that falls outside of those bounds should be investigated. Remember that compromised credentials will look legitimate until you isolate anomalous activity such as moving abnormally large amounts of data, repeated logins during nonbusiness hours or remote access from unusual and faraway locations.
Agencies should be able to identify the following activities:
- Unauthorized access
- Violation of organization policies
- Internal reconnaissance
- Data hoarding
- Data loss
Data analytics can make a huge difference here. If an organization is large, it can be impossible to monitor network activity manually. Anything important is quickly drowned out by the plethora of ordinary information. Using NetFlow and other network metadata, a good security analytics tool can help the relevant information rise to the top.
Agencies also should keep an audit trail of network transactions for as long as is feasible. When struck by an insider attack, the audit trail can be used to identify how the threat operated and what assets were compromised. It may also help the authorities pursue criminal charges against the attacker.
Lastly, don't forget that these attackers thrive on compromising the human layer. Train employees on best practices for using the internet, and on how to recognize social engineering tactics like phishing. Use network segmentation to limit the amount of sensitive data each user has access to, and monitor traffic from third-party contractors for possible compromised credentials.
As the number of access points and connected devices continue to grow, it has become easier for cybercriminals to appear as a legitimate insider. While this trend comes with a new set of challenges than other security concerns, organizations can protect themselves with the right tools and mindset. And early detection of these intruders can keep a security event from becoming the next big breach plastered across the evening news.
NEXT STORY: New bill looks to NIST, FTC for cyber hygiene