{
 "license": "CC-BY-4.0",
 "attribution": "Fide AI, Agentic Cyber Explorer",
 "url": "https://agentic-cyber-explorer.pages.dev/events/glasswing-independent-disclosure-outcomes-study-2026/",
 "asOf": "2026-11-17",
 "id": "glasswing-independent-disclosure-outcomes-study-2026",
 "date": "2026-11-04",
 "datePrecision": "day",
 "title": "Observational Glasswing study reports more disclosures with limited outcome visibility",
 "lane": "defense",
 "kind": "paper",
 "summary": "Thomas Mosk’s observational preprint estimates more than doubled public vulnerability disclosures for Glasswing-access firms relative to comparison firms. Measured remediation did not slow among fixed-and-disclosed flaws. A suggestive decline in the exploited share and no statistically detectable count change do not establish fewer attacks; non-random access and limited disclosure visibility constrain the conclusions.",
 "whyItMatters": "It adds an independent attempt to measure security outcomes while exposing what public disclosure data cannot establish about unseen backlogs or attacks.",
 "actors": [
  "queen-mary-university-of-london"
 ],
 "topics": [
  "vulnerability-discovery",
  "vulnerability-repair",
  "eval-validity",
  "access-controls"
 ],
 "atlas": [],
 "artifacts": [
  "project-glasswing"
 ],
 "sources": [
  {
   "url": "https://arxiv.org/abs/2611.06629v1",
   "publisher": "Thomas Mosk / arXiv",
   "title": "Does Frontier AI Strengthen Cyber Defense? Evidence from Project Glasswing",
   "date": "2026-11-04",
   "type": "primary",
   "accessed": "2026-11-15"
  }
 ],
 "keyFacts": [
  {
   "fact": "The study compares selected Glasswing-access firms with firms without access, including Chinese and Russian firms, using public vulnerability disclosures. Access was not randomly assigned.",
   "locator": "Abstract; empirical design"
  },
  {
   "fact": "The author estimates more than twice as many disclosures and about 150% more high-impact vulnerabilities for access firms relative to comparison firms. These are observational estimates, not a count of flaws independently proven to have been found by AI.",
   "locator": "Abstract; results"
  },
  {
   "fact": "Measured vendor and distribution remediation did not slow among fixed-and-disclosed flaws. This selected sample cannot exclude a growing undisclosed backlog or establish when users installed fixes.",
   "locator": "Section 4: Remediation: was there a human bottleneck?"
  },
  {
   "fact": "The author estimates a roughly two-thirds decline in the share exploited within one month against the Chinese comparison, but calls the decline suggestive. Exploited-vulnerability counts show no statistically detectable change; imprecise estimates allow moderate changes in either direction. This does not establish fewer attacks.",
   "locator": "Section 6: exploitation analysis and inference caveats"
  },
  {
   "fact": "Program selection, national disclosure practices and incomplete vulnerability visibility limit interpretation. Comparisons are between Mythos access and the AI tools available to comparison firms, not AI versus no AI; additional comparison groups and robustness checks do not make access random.",
   "locator": "Introduction limitations; identification and robustness"
  }
 ],
 "significance": 4,
 "fideQuestions": [
  "FID-075",
  "FID-088"
 ],
 "methods": [
  "ai-vulnerability-discovery"
 ],
 "review": "assistant-drafted",
 "addedOn": "2026-11-15",
 "revisions": []
}