Researchers are cloud security firm Aqua have observed threat actors exploiting the recently disclosed Linux privilege escalation flaw Looney Tunables in attacks against cloud environments.
The vulnerability Looney Tunables (CVE-2023-4911 (CVSS score 7.8)) is a buffer overflow issue that resides in the GNU C Library’s dynamic loader ld.so while processing the GLIBC_TUNABLES environment variable. An attacker can trigger the vulnerability to execute code with elevated privileges.
“A buffer overflow was discovered in the GNU C Library’s dynamic loader ld.so while processing the GLIBC_TUNABLES environment variable. This issue could allow a local attacker to use maliciously crafted GLIBC_TUNABLES environment variables when launching binaries with SUID permission to execute code with elevated privileges.” reads the advisory.
Researchers from Aqua Nautilus spotted experimental incursions into cloud environments by Kinsing actors. The attackers used a rudimentary yet PHPUnit vulnerability exploit, they attempted to manipulate the Looney Tunables vulnerability (CVE-2023-4911). The experts pointed out that this marks the first documented instance of such an exploit.
“This marks the first documented instance of such an exploit, to the best of our knowledge.” reads the analysis published by Aqua firm. “Intriguingly, the attacker is also broadening the horizons of their cloud-native attacks by extracting credentials from the Cloud Service Provider (CSP).”
The discovery demonstrates that Kinsing actors are rapidly adding new exploits to their arsenal, expanding the potential targets. Recently, Kinsing actors were observed exploiting vulnerable Openfire servers.
Kinsing actors often exploited the PHPUnit vulnerability (CVE-2017-9841) and it engaged in fully automated attacks as part of mining cryptocurrency. With the recent discovery, the experts observed Kinsing conducting manual tests, which marks a deviation from their usual modus operandi.
The attackers were observed using a Python-based Linux local privilege escalation exploit published by the researcher bl4sty.
“In this particular attack, Kinsing proceeds to retrieve an exploit directly from @bl4sty website. On his site, @bl4sty elucidates that the exploit is a ‘Linux local privilege escalation exploit’ targeting the ‘Looney Tunables’ vulnerability (CVE-2023-4911) found in GNU libc’s ld.so.” continues the analysis.
“He further clarifies that the exploit is grounded in the exploitation methodology detailed in the Qualys writeup, asserting its compatibility with x86(_64) and aarch64 architectures, and highlighting its potential for extension through the addition of new target offsets. The script is accessible for review here.
Subsequently, Kinsing fetches and executes an additional PHP exploit.”
Kinsing threat actors attempt to enumerate the details and credentials associated with the Cloud Service Provider (CSP) to carry out further malicious actions.
The types of credentials and sensitive data that can be compromised by the attackers are Temporary Security Credentials, IAM Role Credentials, and Instance Identity Tokens.
“This marks the inaugural instance of Kinsing actively seeking to gather such information,” security researcher Assaf Morag said.
“The threat actor manually exploited the Looney Tunables vulnerability (CVE-2023-4911) in GNU libc’s ld.so, marking the first known instance of such an exploit. Additionally, Kinsing is expanding its operations by attempting to collect credentials from Cloud Service Providers (CSPs), indicating a potential broadening of their operational scope and an increased threat to cloud-native environments.” concludes the report that also includes Indicators of Compromise (IoCs).
(SecurityAffairs – hacking, Looney Tunables)