CVE-2025-3248
Langflow < 1.3.0 Unauthenticated RCE via /api/v1/validate/code
Exploited in the wild. In CISA KEV since 2025‑05‑05. Fixed versions are out.
Public exploit: 2 sourcesFix4 products2 fixed builds listed below
e.g. Langflow AI - Unauthenticated Remote Code Execution · OID 1.3.6.1.4.1.25623.1.0.154431
What to do
Find your product and version, then upgrade to the build on the right.
- Langflow-ai LangflowAffected: < 1.3.0Upgrade to: 1.3.0Stated by: CVE record
- PyPI langflowAffected: < 1.3.0Upgrade to: 1.3.0Stated by: GitHub advisory
- PyPI langflow-baseAffected: < 0.3.0Upgrade to: 0.3.0Stated by: GitHub advisory
- Сообщество свободного программного обеспечения LangflowAffected: до 1.3.0Upgrade to: not statedStated by: BDU
Versions as the sources state them. Full version matrix after sign-in
Steps
Written by AI from the record- Check whether you run Langflow and what version you have installed (including any langflow-base dependency).
- If your Langflow version is earlier than 1.3.0, plan an immediate upgrade to Langflow 1.3.0.
- If you use/ship langflow-base, upgrade langflow-base to 0.3.0 or later.
- If you cannot upgrade right away, restrict network access so the /api/v1/validate/code endpoint is not reachable from the public internet, and follow your vendor’s mitigation guidance until upgrades are completed.
Coverage timeline
8 publishersAvalon malware framework combines credential collection, lateral movement, and ransomware capabilities.
Microsoft disclosed that threat actor JADEPUFFER used compromised Azure service principals to perform destructive operations against cloud resources.
Advisory coverage
Added to CISA KEV catalog
Sourced from public security reporting — each date is when it was published, not when it happened. 7 further articles named this CVE in wider roundups without reporting it as a distinct event.
What it is
From the CVE record
Langflow versions prior to 1.3.0 are susceptible to code injection in the /api/v1/validate/code endpoint. A remote and unauthenticated attacker can send crafted HTTP requests to execute arbitrary code.
In plain language
Written by AI from the recordLangflow versions before 1.3.0 allow an unauthenticated attacker to run arbitrary code over the internet via /api/v1/validate/code, so most small businesses should treat this as an urgent fix.
Unauthenticated RCE in Langflow via the /api/v1/validate/code endpoint (code validation accepts attacker-controlled input that can lead to arbitrary code execution); this is confirmed in the CISA KEV catalog and is fixed in langflow 1.3.0 (and langflow-base 0.3.0).
If you're affected
- Full server compromise
- Data theft
- Ransomware risk
- Service outage
Exploitation
Where each signal puts this CVE on the scale from published to confirmed exploited.
- CISA KEV
Listed as exploited in the wild since 2025-05-05.
US federal agencies must remediate by 2025-05-26.
Known use in ransomware campaigns.
Apply mitigations per vendor instructions, follow applicable BOD 22-01 guidance for cloud services, or discontinue use of the product if mitigations are unavailable.
- Public exploits
2 sources with a proof of concept or module.
Exploit links, PoCs and Metasploit modules after sign-in- EPSS
100% chance of exploitation activity in the next 30 days, which ranks it in the 99th percentile of scored CVEs.
Exploit Prediction Scoring System, FIRST.org. A probability, not a confirmation.
- Attention now
Sustained, driven by wide press coverage.
Lifecycle
The patch came 120 days before any public exploit.
- Analysis publishedGoogle GTIG Measured the AI Bug Boom: 10,740 Disclosures a Month, AI-Found Flaws Skew RCE — and Only 1 in 431 Gets Exploited
- Analysis publishedMicrosoft Says an Azure Attack Was "Agentic-Driven." Its Own Evidence Says "Scripted."
- Patch availablerecord updated
- Analysis publishedEvery AI Product on CISA's KEV List: All Fourteen of Them
- Analysis publishedIBM Filed Eight More Langflow CVEs Yesterday. That's 111 in Five Months.
- Patch availablerecord updated
Affected products
Technical detail
CVSS 3.1 vector
Open in the CVSS calculatorCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Scored 9.8 by NVD.
How it is reached
- Attack Vector NetworkExploitable remotely over the network without any special conditions
- Attack Complexity LowNo special conditions — the attack can be reliably reproduced
- Privileges Required NoneNo authentication required — anyone can exploit this
- User Interaction NoneNo user interaction needed — fully automated exploitation
Scope
- Scope UnchangedImpact is limited to the vulnerable component itself
Impact if exploited
- Confidentiality HighTotal information disclosure — all data in the component is compromised
- Integrity HighTotal loss of integrity — attacker can modify any data in the component
- Availability HighTotal denial of service — the component is completely unavailable
Weaknesses
ATT&CK techniques
Mapped from the weaknesses above (CWE to ATT&CK), not observed in attacks.
Sources
References in the record
- github.com/langflow-ai/langflow/pull/6911
- github.com/langflow-ai/langflow/releases/tag/1.3.0
- horizon3.ai/attack-research/disclosures/unsafe-at-any-speed-abusing-python-exec-for-unauth-rce-in-langflow-ai/
And 12 more references. See all after sign-in
Analysis on the cve.tools blog
All articles- Google GTIG Measured the AI Bug Boom: 10,740 Disclosures a Month, AI-Found Flaws Skew RCE — and Only 1 in 431 Gets ExploitedOn September 30, Google's Threat Intelligence Group (GTIG) published the first large-scale dataset trying to settle an argument the industry has mostly had with anecdotes: is AI actually accelerating…
- Microsoft Says an Azure Attack Was "Agentic-Driven." Its Own Evidence Says "Scripted."On September 25, 2026, Microsoft's threat intelligence team published a post titled "Storm-3168: Agentic-driven cloud attacks using compromised service principals." The headline attributes a…
- Every AI Product on CISA's KEV List: All Fourteen of ThemFourteen AI-stack CVEs have ever made CISA's KEV catalog, six of them in Langflow alone. Here is the full ledger with dates, the time each bug took to get there, and the 134 AI-stack CVEs that have exploit evidence but no KEV entry at all.
- IBM Filed Eight More Langflow CVEs Yesterday. That's 111 in Five Months.On August 28, IBM's PSIRT filed eight new CVEs against Langflow OSS, the open-source visual builder for AI agents and LLM pipelines that IBM now maintains. The worst of the batch, CVE-2026-19295,…
- CVE-2025-3248: the Langflow endpoint that 'validates' your code by running itCVE-2025-3248 lets anyone who can merely reach a Langflow server execute code on it with no credentials. The bug lives in POST /api/v1/validate/code -- an endpoint whose entire job is to check…
In the news
All news- Amazon Uncovers AI Agent Campaign Targeting LiteLLM, Dify, and Other Apps for API Keys
- The vulnerabilities AI finds are the ones attackers want
- JADEPUFFER-Linked Attackers Used Compromised Service Principals to Delete Azure Resources
- Critical Langflow Flaw Exploited as Attacks on AI Platform Rise
- Hackers Start Exploiting Critical Langflow Vulnerability
- Attackers Exploit Critical Langflow and Rails Flaws in Credential-Probing and C2 Activity
- CISA orders urgent action on actively exploited Langflow RCE flaw
- JadePuffer returns with ransomware built to target AI models and infrastructure
- New ENCFORGE Ransomware Targets AI Model Files in Langflow RCE Attack
- JadePuffer agentic attacks now target AI model data with ransomware
Watch the software you run.
My Stack ranks new CVEs for your products by real-world exploitation, so the next exploited one reaches you without reading every advisory.
We'll flag the next CVE, public exploit or patch for Langflow, not every advisory. This one: actively exploited.
A free account adds
- The full version matrix and every affected product
- Exploit links, proofs of concept and Metasploit modules
- Email alerts for the products you watch
- The same data over REST API, MCP and CLI