A new report shows companies that rely solely on the Common Vulnerabilities and Exposures (CVE) system for their vulnerability information are leaving themselves exposed to a substantial number of security issues they don't know about.
Risk Based Security's researchers have so far this year identified 5,970 more vulnerabilities than reported in the CVE and National Vulnerability Database (NVD). Of them, 18.4% had a CVSS v2 score ranging from 9 to 10, meaning they were considered critical. When vulnerabilities with a severity rating of 7 to 9 were also counted, some 43.5% of the 5,970 flaws not reported in the CVE/NVD system were either high risk or critical. Flaws not listed in CVE/NVD included those involving products from major vendors including Oracle, Microsoft, and Google.
"Organizations that rely on vulnerability intelligence are dealing with an alarming number of issues that impact all parts of their infrastructure," says Brian Martin, vice president of vulnerability intelligence at Risk Based Security.
CVE and NVD only include vulnerabilities that security vendors and researchers directly report to them. As a result, thousands of flaws that researchers discover and disclose in other ways are not getting listed in CVE/NVD, he says. According to Risk Based Security, organizations that rely solely on CVE/NVD likely miss 33% of all disclosed vulnerabilities, on average.
Researchers can disclose vulnerabilities in different ways and different places — from their own blogs to one of millions of repositories on GitHub. "GitHub currently has over 100 million repositories, and that is just a single site," Martin notes. "Factoring in BitBucket, SourceForge, GitLab, and many others, the amount of places a vulnerability may pop up is insane."
Researchers might blog about a vulnerability discovery but often don't cross-post the disclosure to known vulnerability reporting sites such as Bugtraq, Full-Disclosure, or PacketStorm. "Every week we find around a dozen more sources like this, as well as new software being released, or software and third-party libraries," Martin says.
In total, Risk Based Security's VulnDB team aggregated 16,738 security vulnerabilities in the first three quarters of 2019. Nearly half — 48% — had a severity score ranging from 6 to 10. Troublingly, exploit code or proof-of-concept code was available for 39% of the disclosed flaws that Risk Based Security's researchers counted for that time period.
Oracle topped the list of organizations with the most reported security vulnerabilities. Risk Based Security's data showed the company reported 969 security flaws in its products between Jan. 1 and the end of September. Google, with 945 flaws, and SUSE, with 812 flaws, were in second and third place, respectively.
Both Oracle and Google moved up in Risk Based Security's Top 10 list; last year Oracle ranked third and Google ranked fourth among companies that disclosed the most security vulnerabilities in their products. SUSE improved from topping the list last year to moving to the third spot so far in 2019. Meanwhile, Microsoft, which often is perceived as reporting more vulnerabilities than others, was in ninth spot with 485 vulnerabilities, while Cisco landed in 10th spot with 390 vulnerabilities.
As with last year, a majority of the disclosed bugs (10,868) over the past three quarters impacted system integrity. About 1,800 impacted availability and slightly more than 2,600 were related to confidentiality.
- GitHub Becomes CVE Numbering Authority, Acquires Semmle
- MITRE Releases 2019 List of Top 25 Software Weaknesses
- Predicting Vulnerability Weaponization
- 8 Trends in Vulnerability and Patch Management
Check out The Edge, Dark Reading's new section for features, threat data, and in-depth perspectives. Today's top story: "Home Safe: 20 Cybersecurity Tips for Your Remote Workers."