GCVE-1988-2026-0546

Vulnerability from gna-1988 – Published: 2026-11-02 10:19 – Updated: 2026-11-02 10:19
VLAI
Title
[Bugtraq] Bugtraq AI is live
Summary
This is bugtraq@bugtraq.ai, a new mailing list dedicated to the security of AI systems and the use of AI in security research. Bugtraq AI is a sibling of the original Bugtraq list at securityfocus.com, which we are relaunching today after years of silence. Same principles - full disclosure, researcher-first, no corporate filter - applied to a domain that didn't exist when Bugtraq was born in 1993 but now touches everything. What belongs here: - Vulnerabilities in AI/ML systems: model poisoning, training data extraction, prompt injection, inference attacks, jailbreaks with security impact, supply chain issues in model distribution. - Offensive and defensive uses of AI in vulnerability research: automated fuzzing, binary analysis, exploit generation, AI-assisted code audit. - Security implications of AI deployment: model serving infrastructure, API attack surfaces, embedded model security, agent frameworks. - Tooling, techniques, papers, and war stories. What doesn't belong here: hype, product launches, debates without a technical component, or anything you'd post on LinkedIn. AI security is moving fast enough that the traditional conference cycle can't keep up. Papers take months. Blog posts disappear into algorithmic feeds. A mailing list is a permanent, searchable, citable record owned by the community - not by a platform that can change its terms tomorrow. There is a deeper reason this list exists. We are entering an era where AI can generate plausible code, plausible advisories, and plausible explanations of vulnerabilities - all of which may be wrong. Disinformation is no longer just a social media problem; it is a technical one. When a model hallucinates a CVE, or generates a patch that looks correct but isn't, or produces an analysis that passes peer review but not a proof checker - we have a truth problem. Formal methods and mathematical proofs are the antidote. A proof doesn't care who generated it or how convincing the prose around it sounds. It either holds or it doesn't. As AI systems become both the subject and the tool of security research, the ability to formally verify claims - about code, about vulnerabilities, about the models themselves - becomes not academic but essential. This list welcomes that work. If you're applying formal verification to AI systems, using proof assistants to validate security properties, or building tooling that replaces trust with proof - this is the place. The format is old. The problem space is new. That's the point. Post something.
Severity
No CVSS data available.
Assigner
VULNARCHIVE GNA GNA-1988
GNA scorecard E 38/100 over 581 records in the last 180 days details
References
Impacted products
Vendor Product Version CPE status
unknown Bugtraq AI is Affected: unknown
guessed Create a notification for this product.

{
  "containers": {
    "cna": {
      "affected": [
        {
          "product": "Bugtraq AI is",
          "vendor": "unknown",
          "versions": [
            {
              "status": "affected",
              "version": "unknown"
            }
          ]
        }
      ],
      "credits": [
        {
          "lang": "en",
          "type": "finder",
          "value": "Jonathan Brossard"
        }
      ],
      "descriptions": [
        {
          "lang": "en",
          "value": "This is bugtraq@bugtraq.ai, a new mailing list dedicated to the security\nof AI systems and the use of AI in security research.\nBugtraq AI is a sibling of the original Bugtraq list at\nsecurityfocus.com, which we are relaunching today after years of\nsilence. Same principles - full disclosure, researcher-first, no\ncorporate filter - applied to a domain that didn\u0027t exist when Bugtraq\nwas born in 1993 but now touches everything.\nWhat belongs here:\n- Vulnerabilities in AI/ML systems: model poisoning, training data\nextraction, prompt injection, inference attacks, jailbreaks with\nsecurity impact, supply chain issues in model distribution.\n- Offensive and defensive uses of AI in vulnerability research:\nautomated fuzzing, binary analysis, exploit generation, AI-assisted code\naudit.\n- Security implications of AI deployment: model serving infrastructure,\nAPI attack surfaces, embedded model security, agent frameworks.\n- Tooling, techniques, papers, and war stories.\nWhat doesn\u0027t belong here: hype, product launches, debates without a\ntechnical component, or anything you\u0027d post on LinkedIn.\nAI security is moving fast enough that the traditional conference cycle\ncan\u0027t keep up. Papers take months. Blog posts disappear into algorithmic\nfeeds. A mailing list is a permanent, searchable, citable record owned\nby the community - not by a platform that can change its terms tomorrow.\nThere is a deeper reason this list exists. We are entering an era where\nAI can generate plausible code, plausible advisories, and plausible\nexplanations of vulnerabilities - all of which may be wrong.\nDisinformation is no longer just a social media problem; it is a\ntechnical one. When a model hallucinates a CVE, or generates a patch\nthat looks correct but isn\u0027t, or produces an analysis that passes peer\nreview but not a proof checker - we have a truth problem.\nFormal methods and mathematical proofs are the antidote. A proof doesn\u0027t\ncare who generated it or how convincing the prose around it sounds. It\neither holds or it doesn\u0027t. As AI systems become both the subject and\nthe tool of security research, the ability to formally verify claims -\nabout code, about vulnerabilities, about the models themselves - becomes\nnot academic but essential.\nThis list welcomes that work. If you\u0027re applying formal verification to\nAI systems, using proof assistants to validate security properties, or\nbuilding tooling that replaces trust with proof - this is the place.\nThe format is old. The problem space is new. That\u0027s the point.\nPost something."
        }
      ],
      "providerMetadata": {
        "dateUpdated": "2026-11-02T10:19:30Z",
        "orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
        "shortName": "VULNARCHIVE"
      },
      "references": [
        {
          "tags": [
            "technical-description"
          ],
          "url": "https://vuln.freearchive.org/archive/item/9bbf046a7322f59b4b37dcdc1193b54aa518ba700c46ff84776987bddb93af70"
        },
        {
          "tags": [
            "technical-description"
          ],
          "url": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/6ZEX4GCNZEATYIDGO52K37PKMGJCYNAJ/"
        }
      ],
      "source": {
        "defect": [
          "281eaacc-10f4-4abf-8209-231666b27be4%40gmail.com"
        ],
        "discovery": "EXTERNAL"
      },
      "title": "[Bugtraq] Bugtraq AI is live",
      "x_gcve": [
        {
          "recordType": "advisory",
          "relationships": [],
          "vulnId": "GCVE-1988-2026-0546",
          "x_vulnarchive": {
            "archiveUrl": "https://vuln.freearchive.org/archive/item/9bbf046a7322f59b4b37dcdc1193b54aa518ba700c46ff84776987bddb93af70",
            "automated": true,
            "contentSha256": "9bbf046a7322f59b4b37dcdc1193b54aa518ba700c46ff84776987bddb93af70",
            "evidenceScore": 5,
            "messageId": "281eaacc-10f4-4abf-8209-231666b27be4%40gmail.com",
            "originalUrl": "https://lists.securityfocus.com/hyperkitty/list/bugtraq@bugtraq.ai/message/6ZEX4GCNZEATYIDGO52K37PKMGJCYNAJ/",
            "policy": "vulnarchive-1",
            "sourceFormat": "text/html",
            "sourcePublishedAt": "2026-11-02T10:19:30Z"
          }
        }
      ]
    }
  },
  "cveMetadata": {
    "assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
    "assignerShortName": "VULNARCHIVE",
    "datePublished": "2026-11-02T10:19:30Z",
    "dateUpdated": "2026-11-02T10:19:30Z",
    "state": "PUBLISHED",
    "vulnId": "GCVE-1988-2026-0546"
  },
  "dataType": "CVE_RECORD",
  "dataVersion": "5.2"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.


Loading…

Loading…

Loading…

Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.

Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Vulnerability timeline

Sightings and recorded dates, grouped by day · UTC

JSON
How durations are calculated

Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.

Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.

Events included in the timeline

Loading timeline…


    Loading…

    Loading…

    Loading…

    Related by attack behaviour

    Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.