Toward Better Patching — A New Approach with a Dose of AI

Toward Better Patching — A New Approach with a Dose of AI

February 23, 2024 at 09:27AM

Security teams face a daunting task with 2,900 new vulnerabilities predicted monthly in 2024. Effective patching is crucial, yet challenging due to the overwhelming number of vulnerabilities. Coalition’s solution involves a new vulnerability risk prioritization approach using machine learning, aiming to provide timely and reliable vulnerability information while incorporating early threat indicators. This innovative approach shows promise in enhancing vulnerability management efforts.

The main takeaways from the meeting notes are as follows:

1. Security teams are facing a significant challenge in managing the sheer volume of new vulnerabilities, with a prediction of 2,900 new vulnerabilities per month in 2024.

2. Effective patching is a recognized method of reducing breaches, but the high number of known vulnerabilities makes it difficult for security teams to prioritize which ones to address first.

3. Coalition, a cyberinsurance firm, has been researching this issue and aims to help customers improve their patch cycle to reduce claims and increase profits.

4. There are issues with the current methods of vulnerability triaging, including confusing and unreliable sources of information such as media reports and CVSS scores associated with CVE numbers.

5. Coalition has developed a vulnerability risk prioritization solution that utilizes machine learning to generate the Coalition Exploit Scoring System (ESS), providing two scores: the Exploit Availability Probability and the Exploit Usage Probability.

6. Coalition is also working on methods to identify malicious intent in honeypot activity, aiming to enhance accurate and timely vulnerability triaging efforts.

7. The use of AI in vulnerability prioritization approaches is seen as promising, with the potential to cut through the noise and confusion of current methods.

Let me know if I can assist you with anything else.

Full Article