Threat actors have started exploiting a critical-severity remote code execution (RCE) vulnerability in the AI low-code platform Langflow, vulnerability intelligence firm VulnCheck warns.
Tracked as CVE-2026-0768 (CVSS score of 9.8), the security defect exists within the code validator in Langflow’s custom component editor.
Because a user-supplied string is not properly validated before it is used for Python code execution, an attacker could exploit the bug to execute arbitrary code as root without authentication.
The security flaw was reported through ZDI in July 2025 and was publicly disclosed as a zero-day in January 2026. All Langflow releases up to version 1.4.2 are affected.
According to VulnCheck, threat actors have been exploiting the vulnerability to perform reconnaissance and credential harvesting operations.
The cybersecurity firm observed queries for environment variables, secret keys, and SSH access, mainly originating from Russia. By Monday, VulnCheck had seen over 360 exploitation attempts hitting its canaries in the UK.
CVE-2026-0768’s in-the-wild exploitation does not come as a surprise. Last week, VulnCheck warned of a recently observed increase in Langflow vulnerability targeting.
“Before 2026, evidence showed only one Langflow vulnerability known to be exploited in the wild. In 2026, things have changed fast. We’ve now seen 11 additional vulnerabilities targeted and reported as exploited in the wild, highlighting increasing attacker interest in Langflow,” the company said.
VulnCheck has seen more than 15,000 attacks successfully exploiting Langflow instances vulnerable to three of the known exploited flaws, namely CVE-2026-0769, CVE-2025-3248, and CVE-2026-5027.
Related: Critical JFrog Artifactory Vulnerability Reportedly Exploited in the Wild
Related: WatchGuard Patches Critical Vulnerabilities
Related: PaperCut Exploitation Escalates to Active Intrusions
Related: Silent Patches Don’t Stop Attackers – They Blind Defenders
