VULNERABILITY
CVE-2025-59528 in Flowise AI: Critical RCE with Reported Active Exploitation
- A critical remote code execution (RCE) vulnerability, CVE-2025-59528, affects Flowise AI Agent Builder, assigned a CVSS score of 10.0.
- Active exploitation is REPORTED in the wild. The vulnerability has a high EPSS score of 0.8239 (82.39th percentile), indicating a high probability of exploitation.
- Immediate investigation of all publicly exposed Flowise AI instances is required. Restrict network access as a temporary mitigation pending a patch.
Technical Analysis
CVE-2025-59528 is a critical remote code execution vulnerability. While specific technical details of the attack chain are not yet available, the assignment of a CVSS 10.0 score indicates that the vulnerability is likely exploitable over the network with low attack complexity, requires no user interaction, and no privileges on the target system. Successful exploitation results in a complete compromise of the affected host, granting the attacker full control.
Intelligence from The Hacker News specifies the affected product as Flowise AI Agent Builder and reports active RCE exploitation. This contradicts the vendor tag 'Aws' in some initial intelligence feeds, which may be an error or reflect the underlying hosting infrastructure of vulnerable instances. The core threat is an unauthenticated RCE vector in a publicly exposed component of the Flowise AI application.
Affected Systems
- Flowise AI Agent Builder
Intelligence Context
The combination of a perfect CVSS score (10.0), a high EPSS score (0.8239), and credible media reports of active exploitation places CVE-2025-59528 in the highest tier of prioritization. The EPSS score suggests an 82% probability of exploitation activity within the next 30 days, which aligns with the current reports. The vulnerability is not yet listed in the CISA KEV catalog, creating a critical window where attackers are actively exploiting a known vulnerability before it receives wider official alerts. Defenders must act on this intelligence immediately, as automated exploitation is highly probable.
Remediation & Defense
- Restrict all network access to Flowise AI instances to trusted IP addresses immediately.
- If instances must remain exposed, place them behind a web application firewall (WAF) with rulesets designed to block common RCE patterns, though this is not a complete mitigation.
- Monitor for unexpected child processes spawned by the Node.js or application server process associated with Flowise AI.
- Analyze web server and application logs for unusual or malformed API requests that could indicate exploitation attempts.
- Inspect outbound network traffic from Flowise AI servers for connections to anomalous IP addresses or command-and-control infrastructure.
Source Timeline
- nvd.nist.gov — First reporter
- thehackernews.com — Reporting active exploitation