GCVE Launches Decentralized Publishing Ecosystem for Vulnerability Disclosure
The GCVE initiative operated by CIRCL has officially opened its publishing ecosystem, letting organizations issue and share vulnerability identifiers without routing through a central authority.
- Sarah Gooding

GCVE, the decentralized vulnerability identification system operated by Luxembourg's Computer Incident Response Center (CIRCL), launched its publishing ecosystem today alongside the release of Vulnerability-Lookup 4.1.0.
The release comes as the CVE ecosystem's centralization problems are getting harder to ignore. As we covered last week, a surge in OpenClaw advisories put a spotlight on the gaps between GitHub Security Advisories and CVE tracking, and renewed questions about whether the security community's reliance on a single centralized identifier system is still tenable.
GCVE is one concrete answer to that question.
What GCVE Is and How It Works
#GCVE was designed as a backward-compatible extension of the CVE system. Rather than replacing CVE, it introduces a new layer of autonomous publishers called GCVE Numbering Authorities, or GNAs. Each GNA receives a unique numeric identifier that becomes part of the GCVE ID format:
GCVE-<GNA-ID>-<YEAR>-<UNIQUE-ID>
It's basically like the Mastodon of vulnerability identifiers: anyone eligible can run their own instance, issue IDs under their own GNA number, and sync records across the network with no single point of failure or central authority controlling the namespace. The key difference from Mastodon is that GCVE instances aren't siloed by default. Records are designed to propagate across the full network, more like a shared global index than isolated communities. And unlike Mastodon, GCVE isn't a clean break from the existing system. Existing CVEs map into the format under GNA ID 0, so CVE-2023-40224 becomes GCVE-0-2023-40224. CIRCL itself operates as GNA 1.
The design also allows multiple GNAs to publish about the same vulnerability, which Alexandre Dulaunoy, head of security research at CIRCL, described in an August 2025 episode of the Open Source Security podcast as a feature, not a problem. The way he described it, a vendor and an independent researcher can each publish their own advisory about the same issue under different GNA IDs, the same way two developers can maintain forks of the same codebase. A consumer of that data gets both perspectives rather than a single record that tries to reconcile them.
They used Log4Shell as an example. One CVE ID accumulated thousands of references as every vendor, distro, and runtime interpreted the vulnerability differently. As Josh Bressers described it in the podcast: instead of one ID with thousands of references, you get thousands of IDs each with one reference, linked through correlation.
Publishing Is Open to CNAs, CSIRTs, and Vendors With a Disclosure Policy
#The publishing model is built around Vulnerability-Lookup, GCVE's open source reference implementation, which handles record management, metadata enrichment, and synchronization. Organizations run their own instance on their own infrastructure and apply for a GNA ID from gcve.eu. Once assigned, the instance publishes records that other GCVE-compatible systems can discover and sync automatically.
Eligibility covers existing CVE CNAs, registered CSIRTs and CERTs listed at FIRST.org or in the EU CSIRTs Network, and software or hardware vendors with an official CPE vendor name who regularly disclose vulnerabilities in their own products.
Organizations that only want to contribute enrichment data, without running a full advisory workflow, don't need a GNA ID at all. A Vulnerability-Lookup instance is enough to participate in the federation and share metadata like KEV references, threat intelligence context, and cross-references. The GNA directory already includes Red Hat and the European Union Vulnerability Database alongside CIRCL.
Federated Intelligence, Not Just a New ID Format
#In February, GCVE updated db.gcve.eu to run Vulnerability-Lookup 4.0, which introduced federation capabilities that make the ecosystem considerably more useful than a numbering system alone.

With 4.0, independent Vulnerability-Lookup instances can synchronize with each other, sharing structured data including:
- Known Exploited Vulnerabilities (KEV) references
- Sightings
- Analytical bundles
- Community comments and enrichment
Each participating organization controls what it shares. Instances can selectively expose or consume data from trusted peers, which means organizations can contribute context back to the community without giving up control of their infrastructure or data.
Yesterday's 4.1.0 release added full-text search across vulnerability descriptions, CVSS scores, and commit IDs via a new Meilisearch integration, along with eight new vulnerability sources bringing over 96,000 additional advisories into the platform. The new sources span Russian federal vulnerability data (FSTEC BDU), OSV feeds for Haskell, OCaml, AlmaLinux, and OSS-Fuzz, RustSec advisories, Bitnami's container and application database, and CleanStart security advisories. Vulnerability-Lookup now aggregates over 25 sources and more than one million vulnerability records.
CIRCL's own instance already contributes a live stream of vulnerability sightings observed in the wild, which gets shared across the federation as it's collected.
Feedback from the Security Community
#Security researcher Jerry Gamblin, who also contributed to GCVE's BCP-07 KEV standard, weighed in on LinkedIn.
"Interesting update from GCVE today — they've officially launched their publishing ecosystem," Gamblin commented. "The initiative aims to make vulnerability reporting 'simple, independent, and decentralized.' While I love the idea of reducing central bottlenecks, I have a nagging question: is this the evolution we need, or are we heading toward a fragmented landscape where 'global' visibility becomes even harder to maintain?"
Bressers, who has been vocal about CVE's structural problems, commented on Mastodon. "The work the GCVE folks are doing is top notch," he said. "I think they've solved a lot of the problems everyone else is struggling to solve with their federation model."
The CVE system's centralization problems are real: block allocation bottlenecks, policy disputes, coverage gaps, and slow response to surges like the one OpenClaw produced. But decentralization introduces its own coordination costs. If vulnerability identifiers proliferate across independent GNAs with different policies and publication schedules, the tooling that consumes that data has to keep up.
GCVE's answer to this is federation. Rather than requiring a single authoritative registry, the system is designed so that any sufficiently connected instance eventually receives the full picture. That's a reasonable architecture, but it depends on adoption. The more organizations run Vulnerability-Lookup instances and synchronize with the network, the more complete the coverage becomes.
GCVE Aims to Make Vulnerability Data Discoverable Without a Central Gatekeeper
#The CVE Program has faced ongoing criticism over its centralization model, from funding uncertainty to MITRE's role as the administrative backbone to the pace at which advisories from major platforms like GitHub's GHSA get cross-listed as CVEs. Enterprise security tooling, including vulnerability scanners, SBOM tools, and compliance frameworks, are largely built around CVE. Advisories that never get a CVE can be invisible to all of it.
GCVE is trying to build a path where that binary (CVE or invisible) no longer applies. Whether it succeeds depends on whether organizations with meaningful advisory output decide to become GNAs, and whether security tooling vendors start ingesting GCVE data alongside CVE.




