Effective prioritization of vulnerabilities is essential to staying ahead of your attackers. While your threat intelligence might expose a wealth of information about attackers and attack paths, integrating it into decision-making is no easy task. Too often, we make the mistake of taking the data given to us for granted – and this has disastrous consequences. In this blog post, I’ll explain what we miss by trusting CVSS scores, and what should absolutely be taken into consideration to focus on the vulnerabilities posing the greatest risks to our organizations.
Part of what Risk I/O does as a vulnerability threat management provider is leverage threat intelligence sources to help our customers understand the likelihood of a vulnerability breach. While still not a complete picture of the threat landscape, we use data from public vulnerability databases, zero-day vulnerability intelligence and aggregated metadata from Risk I/O’s 10,000 enterprise users, 1,100,000 live assets and over 100 million live vulnerabilities to assess the effectiveness of CVSS as a remediation policy.
Figure 1: Risk I/O correlates attack, threat and exploit data against user vulnerabilities data 24/7 to assess the effectiveness of CVSS as a remediation policy.
And what we’ve found is that some of the most damaging CVE’s have been incorrectly assigned with “low” CVSS scores. What are some types of low CVSS scores that are currently being attacked?
Dell CTU researchers have found significant scanning activity in most all sectors and across the world related to these vulnerabilities, and at Risk I/O we’ve observed over 200 million breaches, yet we’re still stuck basing our remediation policies on CVSS and vendor assigned scores. Why?
Where Is CVSS Failing?
CVSS scoring is failing to take into account quite a few factors:
1. Targets of opportunity for attackers:
The amount of sites affected by CVE-2014-0160 is unfathomable – with broad estimates between 30-70% of the Internet. And randomly selecting vulnerabilities from a stack gives one about a 2% chance of remediating a truly critical vulnerability. All of these vulnerabilities give attackers a known edge on the probability that randomly targeting with a weaponized exploit will yield results – and this is why they use them.
2. Active and successful in-the-wild exploitation:
We are logging about 2M breaches (or successful exploits) every day across live vulnerabilities. The rate of incoming breaches is also increasing.
3. They’re easy to exploit:
Metasploit module and ExploitDB are databases that offer information about attackers’ behaviors. This informs our decision making about vulnerability assessment and remediations. The best policy is fixing vulnerabilities with entries in both Metasploit and ExploitDB, yielding about a 30% success rate (or 9x better than anything CVSS gets to).
What About the Security Risk Assessment Methods Used Today?
Security risk assessment has been lagging behind our capabilities for years. With the constant release of a countless volume of vulnerabilities, we have hard evidence on why and where our vulnerability scoring systems are failing us. Today, we have access to free exploit databases, open threat exchanges, and a number of proprietary tools. With cheap cloud resources available, we no longer need to rely purely on analysts’ opinions of what sort of risk a vulnerability posts. Instead, we can add context through structured, real-time analysis of the data.
But current risk assessment methodologies do not ﬁt real “in-the-wild” attack data. Mauricio Velazco, the head of vulnerability management at the Blackstone Group drives the point home in an article in which he explained, “We have to mitigate risk before the exploit happens. If you try to mitigate after, that is more costly, has more impact, and is more dangerous for your company.” Current prioritization strategies based on the Common Vulnerability Scoring System (CVSS), and subsequent adaptations of such scores, have two fatal flaws:
1. Current risk assessments (CVSS included) lack information about what kind of attacks are possible. As security practitioners, we care about which vulnerabilities matter. We solve this problem by using live, currently open vulnerabilities to do our assessments.
2. Attackers do not go after the same vulnerability month after month, week after week, hour after hour. If certain types of attacks are failing, they change strategies. Below is a timeline of breach counts across 30,000 organizations worldwide. Each color represents a different CVE, and these are only the ones which have more than 10,000 breaches recorded to their name.
Figure 2: This timeline represents the CVEs with more than 10,000 breaches recorded to their name. Note that the type of CVE exploited changes, as well as the frequency with which they are exploited.
Current risk assessment strategies are based on CVSS scores, which are assigned sometimes years before an organization makes the unwise decision to patch or forego that vulnerability. Risks change in real-time, and as a result, risk assessment methodologies should be real-time as well.
What Do We Do About It?
Too often, infosec professionals find themselves working hard at the wrong thing. Working on the right thing is probably more important than working hard. So how should you prioritize the remediation of some vulnerabilities over others?
A better strategy would be to use threat intelligence to explore the source and spread of vulnerabilities. Since breach data stems from how attackers are behaving, having a handle on threat intelligence allows you to identify which vulnerabilities have a high probability of causing a breach. Checking threat intelligence feeds for what’s being actively exploited, to think like an attacker and to have the same information an attacker has is an action plan that infosec professionals can take to prioritize remediation. This shifts your strategy away from trying to fix everything and instead, focusing on identifying and remediating the few vulnerabilities which are most likely to cause a breach.
You can hear Michael Roytman’s discuss vulnerability prioritization at a NY Information Security Meetup.