In the wake of the recent Microsoft Azure vulnerability, ChaosDB, security experts are stressing that organizations, especially those that rely on public cloud infrastructure can no longer delay adopting the zero-trust model.
Cloud security firm Wiz, which first discovered the vulnerability in Microsoft Azure’s managed database service, Cosmos DB said that the vulnerability gave threat actors “complete, unrestricted access to the accounts and databases of several thousand Microsoft Azure companies, including many Fortune 500 companies.”
The flaw was due to a misconfigured new feature, called Jupyter Notebook, that Microsoft added to CosmosDB. Anyone could exploit the misconfiguration to escalate privileges and steal the access keys of other customers’ notebooks and then use it to control Cosmos DB accounts directly over the internet.
The flaw is regarded as one of the most dangerous of all time as it compromised the core tenet of public cloud security.
“One tenant – customer – of a cloud provider should not have the ability to access or affect another tenant without authorization. Secure multitenancy is a core building block of public cloud security, and without it, no organization or individual could trust a public cloud.”, said Nutkis from cloud security firm Okay9 to DataCentre Knowledge.
Microsoft disabled the vulnerable notebooks within two days of it being alerted by Wiz and also said it has found no evidence of the flaw being exploited. But security experts fear if any attacker got access to the keys, they might still be able to access the database.
Since Microsoft couldn’t reset the private security keys by themselves, they reached out to customers and instructed them to perform the reset as a precaution.
Security experts said zero trust and identity-based authentication are two strategies that can protect organizations from flaws like the ChaosDB, that affect the very foundation of public cloud security and are only likely to increase with the increased reliance on cloud infrastructure services.
“Anytime we have these highly visible, high-profile weaknesses, attackers are going to notice that, and it’s going to inspire similar attacks, similar offensive research,” Mark Orlando, co-founder, and CEO at Bionic Cyber told DataCentre Knowledge.
Zero-Trust is crucial because it’s a strategy that limits what the attackers can do even if they get access to the security keys or credentials, as traffic even within the network is treated as unsafe until proven otherwise.
For instance, in the case of the ChaosDB vulnerability, if organizations put in place controls that ensure no access is granted to untrusted IP addresses or untrusted endpoints or from accounts that can’t verify their identity then attackers can’t progress with their objectives even if they have access to the private keys.
Pointing to security measures like zero-trust and identity-based authentication, Orlando said, “even if credentials and keys are compromised, the damage is limited,”. “It’s not a global issue.”
A July survey by Microsoft revealed 96% of security decision-makers acknowledged that Zero-Trust is critical for their organization’s success, 76% are in the process of implementing it and only 35% of them have fully implemented Zero-Trust.
ManageEngine has a webinar that unpacks the National Institute of Standards and Technology’s Zero Trust model, discusses the challenges that organizations are facing with its adoption, and provides action items on how organizations can accelerate zero-trust implementation. You can watch it here.