Vulnerable Apace ActiveMQ instances impacted by the maximum severity bug, tracked as CVE-2023-46604, could be exploited to achieve arbitrary code execution from memory through a new proof-of-concept exploit, The Hacker News reports.
While previous attacks targeted at the flaw used ActiveMQ's ClassPathXmlApplicationContext class to facilitate malicious XML bean configuration file loading and remote code execution, VulnCheck researchers noted that an exploit using the FileSystemXmlApplicationContext class with a custom SpEL expression yielded the same results without placing their tools to disk.
However, such an exploit would prompt an exception message that requires additional measures to ensure the concealment of malicious activity, according to the report.
"Now that we know attackers can execute stealthy attacks using CVE-2023-46604, it's become even more important to patch your ActiveMQ servers and, ideally, remove them from the internet entirely," said VulnCheck Chief Technology Officer Jacob Baines.
Included in the exposed dataset labeled "PDL" were individuals' full names, email addresses, phone numbers, location data, professional summaries, and skills, as well as education and employment histories.
Attackers could leverage the issue — which impacts Nexus 3000, 7000, and 9000 series switches with vulnerable NX-OS versions with DHCPv6 activated and are in standalone NX-OS mode — to facilitate continuous crashes of the dhcp_snoop process and a denial-of-service condition.
Higher rewards of up to $250,000 will be given by Google for the discovery of memory corruption flaws in the Chrome browser shown to achieve remote code execution using a non-sandboxed process as part of a more robust vulnerability reward program.