Microsoft has silently fixed an important-severity security flaw in its Azure Cognitive Search (ACS) after an external researcher warned that a buggy feature allowed cross-tenant network bypass attacks.
The vulnerability, documented by researchers at Mnemonic, effectively removed the entire network and identity perimeter around internet-isolated Azure Cognitive Search instances and allowed cross-tenant access to the data plane of ACS instances from any location, including instances without any explicit network exposure.
According to Mnemonic researcher Emilien Socchi, the flaw was silently fixed by Microsoft at the end of August 2022, approximately six months after it was first reported.
The exposure, nicknamed ACSESSED, impacted all Azure Cognitive Search instances that enabled the “Allow access from portal” feature.
“By enabling that feature, customers effectively allowed cross-tenant access to the data plane of their ACS instances from any location, regardless of the actual network configurations of the latter. Note that this included instances exposed exclusively on private endpoints, as well as instances without any explicit network exposure, such as the one I deployed for investigation (i.e. instances without any private, service or public endpoint),” the researcher warned.
“By the simple click of a button, customers were able to turn on a vulnerable feature, which removed the entire network perimeter configured around their ACS instances, without providing any real identity perimeter (i.e. anybody could generate a valid access token for ARM),” Socchi added.
The Mnemonic researcher said Microsoft paid a $10,000 bounty and elevated the risk level from moderate to important because of the cross-tenant risk and ease of exploitation.
At one point during the disclosure process, Microsoft said the patch was delayed because the fix required “a significant design level change.”
*an earlier version of the article incorrectly mentioned Azure Container Services, which has the same acronym as Azure Cognitive Search, ACS
Related: Attacks Targeting Azure OMIGOD Vulnerability Ramping Up
Related: For Microsoft, Security is a $10 Billion Business
Related: Microsoft Confirms ‘NotLegit’ Azure Flaw Exposed Source Code
Ryan Naraine is Editor-at-Large at SecurityWeek and host of the popular Security Conversations podcast series. He is a security community engagement expert who has built programs at major global brands, including Intel Corp., Bishop Fox and GReAT. Ryan is a founding-director of the Security Tinkerers non-profit, an advisor to early-stage entrepreneurs, and a regular speaker at security conferences around the world.
More from Ryan Naraine
- Skybox Security Raises $50M, Hires New CEO
- OpenSSL Ships Patch for High-Severity Flaws
- Comcast Wants a Slice of the Enterprise Cybersecurity Business
- VMware Confirms Exploit Code Released for Critical vRealize Logging Vulnerabilities
- Gem Security Gets $11 Million Seed Investment for Cloud Incident Response Platform
- Ransomware Leads to Nantucket Public Schools Shutdown
- Sentra Raises $30 Million for DSPM Technology
- Saviynt Raises $205M; Founder Rejoins as CEO
Latest News
- Skybox Security Raises $50M, Hires New CEO
- Spies, Hackers, Informants: How China Snoops on the US
- Australian Man Sentenced for Scam Related to Optus Hack
- Chrome 110 Patches 15 Vulnerabilities
- Application Security Protection for the Masses
- Tor Network Under DDoS Pressure for 7 Months
- Siemens License Manager Vulnerabilities Allow ICS Hacking
- UN Experts: North Korean Hackers Stole Record Virtual Assets
