Cybercriminals can exploit a design flaw in the Microsoft Exchange Autodiscover protocol to “leak” web requests outside of an end-user’s domains, according to the cybersecurity research team at network segmentation solutions provider Guardicore. Autodiscover is a protocol used by Exchange for automatic configuration of Microsoft Outlook and other clients.
The Autodiscover flaw enables cybercriminals to control user domains or “sniff” traffic in the same network and capture domain credentials in plain text (HTTP basic authentication) that are being transferred over the wire, Guardicore said. It also allows hackers with DNS-poisoning capabilities to syphon leaky passwords.
Analyzing the Exchange Autodiscover Flaw
Guardicore acquired multiple Autodiscover domains with a TLD suffix and set them up to reach a web server, the company said. It then detected a leak of Windows domain credentials that reached its server.
In total, Guardicore captured 372,072 Windows domain credentials due to the Autodiscover flaw, the company indicated. Guardicore also obtained 96,671 unique credentials that leaked from mobile email clients and other applications interfacing with the Exchange server.
How to Avoid an Autodiscover Leak: To mitigate the Autodiscover design flaw, users can block Autodiscover domains in their firewalls, Guardicore stated. In addition, they can leverage HTTP basic authentication and verify that support for basic authentication is disabled when they deploy or configure Exchange setups.
Meanwhile, software developers and vendors should verify that they are implementing Autodiscover in their products and not letting them “fail upwards,” Guardicore indicated. This ensures that software developers and vendors should avoid developing domains such as “Autodiscover” with a “back-off” algorithm.
Microsoft Exchange Security Challenges
This is the latest in a growing list of Microsoft Exchange-related security issues that MSSPs and MSPs have had to mitigate for customers.