Sweet Security adds one click virtual patching for cloud apps

Sweet Security, a company in Tel Aviv that protects cloud apps, announced virtual patching. A person sends one policy update to the sensors on the cloud servers. The sensors stop attacks on a known vulnerability. The company says this gives engineers time to make a permanent patch. Mandiant says the mean time to exploit fell to about minus 7 days. This means attacks can come before a patch.

How it works

Sensors sit on the cloud workloads, the servers that operate the apps. They learn the usual behavior of an app and stop behavior that is different. The tool gives each vulnerability a score. A person sends the policy update, and it stops requests that use the vulnerability at the network layer.

The numbers

CNBC reported that Bank of America hired 1,000 contractors to patch software vulnerabilities. The company gave no test results and no customer numbers for the new tool.

What is not known

Analyst James Berthoty says virtual patching is a stopgap that buys time to patch. Four other vendors have built tools that are almost the same. No data shows how good they are.

Insight: Vendors sell time because patching is late

Mandiant puts the mean time to exploit at about minus 7 days. 5 or more vendors sell a stopgap for this. A rule monitor for AI agents found about 70% of attacks in a test, but it gave an alarm for 29.3% of safe runs. This reads as a race to own the layer that works before the patch, and the rate of incorrect alarms can decide who wins.