As per recent media reports, multiple SAP AI Core vulnerabilities have been discovered by cybersecurity researchers. This is a cloud-based platform used for creating and deploying AI workflows. In this article, we’ll look at how these SAP AI Core flaws were discovered and what they could be exploited for. Let’s begin!
SAP AI Core Vulnerabilities: Initial Discovery
Before we get into the details of the vulnerabilities, it’s that SAP AI Core is a platform that operates based on a cloud model. The platform can be used to access essential tools and infrastructure tools for AI workflows. These tools and the infrastructure allow them to construct, deploy, and manage predictive AI workflows.
The SAP AI Core vulnerabilities were initially discovered by cybersecurity researchers at Wiz. The researchers unearthed five SAP security shortcomings that are now being tracked as SAPwned. To identify these vulnerabilities, the researchers focused on security risks associated with AI training services that require access to sensitive data.
As a result of the research, it was discovered that access to customers’ private data and credentials, located across various cloud services, could be acquired by executing legitimate AI training procedures. Such access would allow threat actors to read and modify docker image artifacts and gain administrative privileges on SAP’s Kubernetes cluster.
Consequences of Potential Exploits
According to the information available, the acquired access could be weaponized further since the Helm package manager server was exposed to both read and write operations. Commenting on the SAP AI Core flaws, security researcher Ben-Sasson has stated that:
“Using this access level, an attacker could directly access other customer’s Pods and steal sensitive data, such as models, datasets, and code. This access also allows attackers to interfere with customer’s Pods, taint AI data, and manipulate models’ inference.”
Apart from this, these flaws make it possible for threat actors to gain unauthorized access to data and also to credentials for cloud environments such as Microsoft Azure, Amazon Web Services (AWS), and SAP HANA Cloud.
Causes Behind The Sap AI Core Vulnerabilities
Shedding light on the causes of such vulnerabilities, Wiz claims that the issues surfaced as a result of the platform making it possible to run malicious AI models and training procedures without sandboxing or having sufficient isolation measures in place. In addition, media reports have cited Ben-Sasson stating that:
“The recent security flaws in AI service providers like Hugging Face, Replicate, and SAP AI Core highlight significant vulnerabilities in their tenant isolation and segmentation implementations. These platforms allow users to run untrusted AI models and training procedures in shared environments, increasing the risk of malicious users being able to access other users’ data.”
What these vulnerabilities essentially mean is that AI training requires arbitrary code. In light of such requirements, organizations must have sufficient guardrails in place to lower risk exposure and improve security posture.
Conclusion
The discovery of SAP AI Core flaws emphasizes the critical need for robust security measures. Effective tenant isolation and sandboxing are essential to prevent unauthorized access and protect sensitive data. Addressing such flaws will enhance the security and reliability of AI services in cloud environments.
The sources for this piece include articles in The Hacker News and Security Affairs.
The post Alert: SAP AI Core Vulnerabilities Put Customer Data At Risk appeared first on TuxCare.
*** This is a Security Bloggers Network syndicated blog from TuxCare authored by Wajahat Raja. Read the original post at: https://tuxcare.com/blog/alert-sap-ai-core-vulnerabilities-put-customer-data-at-risk/