[{"data":1,"prerenderedAt":791},["ShallowReactive",2],{"/en-us/blog/categories/security":3,"navigation-en-us":20,"banner-en-us":430,"footer-en-us":440,"security-category-page-total-items-en-us":682,"security-category-page-featured-en-us":683,"security-category-page-1-en-us":712},{"id":4,"title":5,"body":6,"category":6,"config":7,"content":11,"description":6,"extension":12,"meta":13,"navigation":14,"path":15,"seo":16,"slug":6,"stem":18,"testContent":6,"type":6,"__hash__":19},"blogCategories/en-us/blog/categories/security.yml","Security",null,{"template":8,"slug":9,"hide":10},"BlogCategory","security",false,{"name":5},"yml",{},true,"/en-us/blog/categories/security",{"title":5,"description":17},"Browse articles related to Security on the GitLab Blog","en-us/blog/categories/security","Hx58KagneyLDkWgUOsPQNGCsWqekf9YGQa6EJFfGFRw",{"data":21},{"logo":22,"freeTrial":27,"sales":32,"login":37,"items":42,"search":350,"minimal":381,"duo":400,"switchNav":409,"pricingDeployment":420},{"config":23},{"href":24,"dataGaName":25,"dataGaLocation":26},"/","gitlab logo","header",{"text":28,"config":29},"Get free trial",{"href":30,"dataGaName":31,"dataGaLocation":26},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com&glm_content=default-saas-trial/","free trial",{"text":33,"config":34},"Talk to sales",{"href":35,"dataGaName":36,"dataGaLocation":26},"/sales/","sales",{"text":38,"config":39},"Sign in",{"href":40,"dataGaName":41,"dataGaLocation":26},"https://gitlab.com/users/sign_in/","sign in",[43,70,164,169,271,331],{"text":44,"config":45,"cards":47},"Platform",{"dataNavLevelOne":46},"platform",[48,54,62],{"title":44,"description":49,"link":50},"The intelligent orchestration platform for DevSecOps",{"text":51,"config":52},"Explore our Platform",{"href":53,"dataGaName":46,"dataGaLocation":26},"/platform/",{"title":55,"description":56,"link":57},"GitLab Duo Agent Platform","Agentic AI for the entire software lifecycle",{"text":58,"config":59},"Meet GitLab Duo",{"href":60,"dataGaName":61,"dataGaLocation":26},"/gitlab-duo-agent-platform/","gitlab duo agent platform",{"title":63,"description":64,"link":65},"Why GitLab","See the top reasons enterprises choose GitLab",{"text":66,"config":67},"Learn more",{"href":68,"dataGaName":69,"dataGaLocation":26},"/why-gitlab/","why gitlab",{"text":71,"left":14,"config":72,"link":74,"lists":78,"footer":146},"Product",{"dataNavLevelOne":73},"solutions",{"text":75,"config":76},"View all Solutions",{"href":77,"dataGaName":73,"dataGaLocation":26},"/solutions/",[79,103,125],{"title":80,"description":81,"link":82,"items":87},"Automation","CI/CD and automation to accelerate deployment",{"config":83},{"icon":84,"href":85,"dataGaName":86,"dataGaLocation":26},"AutomatedCodeAlt","/solutions/delivery-automation/","automated software delivery",[88,92,95,99],{"text":89,"config":90},"CI/CD",{"href":91,"dataGaLocation":26,"dataGaName":89},"/solutions/continuous-integration/",{"text":55,"config":93},{"href":60,"dataGaLocation":26,"dataGaName":94},"gitlab duo agent platform - product menu",{"text":96,"config":97},"Source Code Management",{"href":98,"dataGaLocation":26,"dataGaName":96},"/solutions/source-code-management/",{"text":100,"config":101},"Automated Software Delivery",{"href":85,"dataGaLocation":26,"dataGaName":102},"Automated software delivery",{"title":5,"description":104,"link":105,"items":110},"Deliver code faster without compromising security",{"config":106},{"href":107,"dataGaName":108,"dataGaLocation":26,"icon":109},"/solutions/application-security-testing/","security and compliance","ShieldCheckLight",[111,115,120],{"text":112,"config":113},"Application Security Testing",{"href":107,"dataGaName":114,"dataGaLocation":26},"Application security testing",{"text":116,"config":117},"Software Supply Chain Security",{"href":118,"dataGaLocation":26,"dataGaName":119},"/solutions/supply-chain/","Software supply chain security",{"text":121,"config":122},"Software Compliance",{"href":123,"dataGaName":124,"dataGaLocation":26},"/solutions/software-compliance/","software compliance",{"title":126,"link":127,"items":132},"Measurement",{"config":128},{"icon":129,"href":130,"dataGaName":131,"dataGaLocation":26},"DigitalTransformation","/solutions/visibility-measurement/","visibility and measurement",[133,137,141],{"text":134,"config":135},"Visibility & Measurement",{"href":130,"dataGaLocation":26,"dataGaName":136},"Visibility and Measurement",{"text":138,"config":139},"Value Stream Management",{"href":140,"dataGaLocation":26,"dataGaName":138},"/solutions/value-stream-management/",{"text":142,"config":143},"Analytics & Insights",{"href":144,"dataGaLocation":26,"dataGaName":145},"/solutions/analytics-and-insights/","Analytics and insights",{"title":147,"items":148},"GitLab for",[149,154,159],{"text":150,"config":151},"Enterprise",{"href":152,"dataGaLocation":26,"dataGaName":153},"/enterprise/","enterprise",{"text":155,"config":156},"Small Business",{"href":157,"dataGaLocation":26,"dataGaName":158},"/small-business/","small business",{"text":160,"config":161},"Public Sector",{"href":162,"dataGaLocation":26,"dataGaName":163},"/solutions/public-sector/","public sector",{"text":165,"config":166},"Pricing",{"href":167,"dataGaName":168,"dataGaLocation":26,"dataNavLevelOne":168},"/pricing/","pricing",{"text":170,"config":171,"link":173,"lists":177,"feature":262},"Resources",{"dataNavLevelOne":172},"resources",{"text":174,"config":175},"View all resources",{"href":176,"dataGaName":172,"dataGaLocation":26},"/resources/",[178,211,234],{"title":179,"items":180},"Getting started",[181,186,191,196,201,206],{"text":182,"config":183},"Install",{"href":184,"dataGaName":185,"dataGaLocation":26},"/install/","install",{"text":187,"config":188},"Quick start guides",{"href":189,"dataGaName":190,"dataGaLocation":26},"/get-started/","quick setup checklists",{"text":192,"config":193},"Learn",{"href":194,"dataGaLocation":26,"dataGaName":195},"https://university.gitlab.com/","learn",{"text":197,"config":198},"Product documentation",{"href":199,"dataGaName":200,"dataGaLocation":26},"https://docs.gitlab.com/","product documentation",{"text":202,"config":203},"Best practice videos",{"href":204,"dataGaName":205,"dataGaLocation":26},"/getting-started-videos/","best practice videos",{"text":207,"config":208},"Integrations",{"href":209,"dataGaName":210,"dataGaLocation":26},"/integrations/","integrations",{"title":212,"items":213},"Discover",[214,219,224,229],{"text":215,"config":216},"Customer success stories",{"href":217,"dataGaName":218,"dataGaLocation":26},"/customers/","customer success stories",{"text":220,"config":221},"Blog",{"href":222,"dataGaName":223,"dataGaLocation":26},"/blog/","blog",{"text":225,"config":226},"The Source",{"href":227,"dataGaName":228,"dataGaLocation":26},"/the-source/","the source",{"text":230,"config":231},"Remote",{"href":232,"dataGaName":233,"dataGaLocation":26},"https://handbook.gitlab.com/handbook/company/culture/all-remote/","remote",{"title":235,"items":236},"Connect",[237,242,247,252,257],{"text":238,"config":239},"GitLab Services",{"href":240,"dataGaName":241,"dataGaLocation":26},"/services/","services",{"text":243,"config":244},"Community",{"href":245,"dataGaName":246,"dataGaLocation":26},"/community/","community",{"text":248,"config":249},"Forum",{"href":250,"dataGaName":251,"dataGaLocation":26},"https://forum.gitlab.com/","forum",{"text":253,"config":254},"Events",{"href":255,"dataGaName":256,"dataGaLocation":26},"/events/","events",{"text":258,"config":259},"Partners",{"href":260,"dataGaName":261,"dataGaLocation":26},"/partners/","partners",{"textColor":263,"title":264,"text":265,"link":266},"#000","What’s new in GitLab","Stay updated with our latest features and improvements.",{"text":267,"config":268},"Read the latest",{"href":269,"dataGaName":270,"dataGaLocation":26},"/releases/whats-new/","whats new",{"text":272,"config":273,"lists":275},"Company",{"dataNavLevelOne":274},"company",[276],{"items":277},[278,283,289,291,296,301,306,311,316,321,326],{"text":279,"config":280},"About",{"href":281,"dataGaName":282,"dataGaLocation":26},"/company/","about",{"text":284,"config":285,"footerGa":288},"Jobs",{"href":286,"dataGaName":287,"dataGaLocation":26},"/jobs/","jobs",{"dataGaName":287},{"text":253,"config":290},{"href":255,"dataGaName":256,"dataGaLocation":26},{"text":292,"config":293},"Leadership",{"href":294,"dataGaName":295,"dataGaLocation":26},"/company/team/e-group/","leadership",{"text":297,"config":298},"Team",{"href":299,"dataGaName":300,"dataGaLocation":26},"/company/team/","team",{"text":302,"config":303},"Handbook",{"href":304,"dataGaName":305,"dataGaLocation":26},"https://handbook.gitlab.com/","handbook",{"text":307,"config":308},"Investor relations",{"href":309,"dataGaName":310,"dataGaLocation":26},"https://ir.gitlab.com/","investor relations",{"text":312,"config":313},"Trust Center",{"href":314,"dataGaName":315,"dataGaLocation":26},"/security/","trust center",{"text":317,"config":318},"AI Transparency Center",{"href":319,"dataGaName":320,"dataGaLocation":26},"/ai-transparency-center/","ai transparency center",{"text":322,"config":323},"Newsletter",{"href":324,"dataGaName":325,"dataGaLocation":26},"/company/contact/#contact-forms","newsletter",{"text":327,"config":328},"Press",{"href":329,"dataGaName":330,"dataGaLocation":26},"/press/","press",{"text":332,"config":333,"lists":334},"Contact us",{"dataNavLevelOne":274},[335],{"items":336},[337,340,345],{"text":33,"config":338},{"href":35,"dataGaName":339,"dataGaLocation":26},"talk to sales",{"text":341,"config":342},"Support portal",{"href":343,"dataGaName":344,"dataGaLocation":26},"https://support.gitlab.com","support portal",{"text":346,"config":347},"Customer portal",{"href":348,"dataGaName":349,"dataGaLocation":26},"https://customers.gitlab.com/customers/sign_in/","customer portal",{"close":351,"login":352,"suggestions":359},"Close",{"text":353,"link":354},"To search repositories and projects, login to",{"text":355,"config":356},"gitlab.com",{"href":40,"dataGaName":357,"dataGaLocation":358},"search login","search",{"text":360,"default":361},"Suggestions",[362,364,368,370,374,378],{"text":55,"config":363},{"href":60,"dataGaName":55,"dataGaLocation":358},{"text":365,"config":366},"Code Suggestions (AI)",{"href":367,"dataGaName":365,"dataGaLocation":358},"/solutions/code-suggestions/",{"text":89,"config":369},{"href":91,"dataGaName":89,"dataGaLocation":358},{"text":371,"config":372},"GitLab on AWS",{"href":373,"dataGaName":371,"dataGaLocation":358},"/partners/technology-partners/aws/",{"text":375,"config":376},"GitLab on Google Cloud",{"href":377,"dataGaName":375,"dataGaLocation":358},"/partners/technology-partners/google-cloud-platform/",{"text":379,"config":380},"Why GitLab?",{"href":68,"dataGaName":379,"dataGaLocation":358},{"freeTrial":382,"mobileIcon":387,"desktopIcon":392,"secondaryButton":395},{"text":383,"config":384},"Start free trial",{"href":385,"dataGaName":31,"dataGaLocation":386},"https://gitlab.com/-/trials/new/","nav",{"altText":388,"config":389},"Gitlab Icon",{"src":390,"dataGaName":391,"dataGaLocation":386},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1758203874/jypbw1jx72aexsoohd7x.svg","gitlab icon",{"altText":388,"config":393},{"src":394,"dataGaName":391,"dataGaLocation":386},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1758203875/gs4c8p8opsgvflgkswz9.svg",{"text":396,"config":397},"Get Started",{"href":398,"dataGaName":399,"dataGaLocation":386},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com/get-started/","get started",{"freeTrial":401,"mobileIcon":405,"desktopIcon":407},{"text":402,"config":403},"Learn more about GitLab Duo",{"href":60,"dataGaName":404,"dataGaLocation":386},"gitlab duo",{"altText":388,"config":406},{"src":390,"dataGaName":391,"dataGaLocation":386},{"altText":388,"config":408},{"src":394,"dataGaName":391,"dataGaLocation":386},{"button":410,"mobileIcon":415,"desktopIcon":417},{"text":411,"config":412},"/switch",{"href":413,"dataGaName":414,"dataGaLocation":386},"#contact","switch",{"altText":388,"config":416},{"src":390,"dataGaName":391,"dataGaLocation":386},{"altText":388,"config":418},{"src":419,"dataGaName":391,"dataGaLocation":386},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1773335277/ohhpiuoxoldryzrnhfrh.png",{"freeTrial":421,"mobileIcon":426,"desktopIcon":428},{"text":422,"config":423},"Back to pricing",{"href":167,"dataGaName":424,"dataGaLocation":386,"icon":425},"back to pricing","GoBack",{"altText":388,"config":427},{"src":390,"dataGaName":391,"dataGaLocation":386},{"altText":388,"config":429},{"src":394,"dataGaName":391,"dataGaLocation":386},{"title":431,"button":432,"config":437},"See how agentic AI transforms software delivery",{"text":433,"config":434},"Watch GitLab Transcend now",{"href":435,"dataGaName":436,"dataGaLocation":26},"/events/transcend/virtual/","transcend event",{"layout":438,"icon":439,"disabled":14},"release","AiStar",{"data":441},{"text":442,"source":443,"edit":449,"contribute":454,"config":459,"items":464,"minimal":671},"Git is a trademark of Software Freedom Conservancy and our use of 'GitLab' is under license",{"text":444,"config":445},"View page source",{"href":446,"dataGaName":447,"dataGaLocation":448},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/","page source","footer",{"text":450,"config":451},"Edit this page",{"href":452,"dataGaName":453,"dataGaLocation":448},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/content/","web ide",{"text":455,"config":456},"Please contribute",{"href":457,"dataGaName":458,"dataGaLocation":448},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/CONTRIBUTING.md/","please contribute",{"twitter":460,"facebook":461,"youtube":462,"linkedin":463},"https://twitter.com/gitlab","https://www.facebook.com/gitlab","https://www.youtube.com/channel/UCnMGQ8QHMAnVIsI3xJrihhg","https://www.linkedin.com/company/gitlab-com",[465,512,566,610,637],{"title":165,"links":466,"subMenu":481},[467,471,476],{"text":468,"config":469},"View plans",{"href":167,"dataGaName":470,"dataGaLocation":448},"view plans",{"text":472,"config":473},"Why Premium?",{"href":474,"dataGaName":475,"dataGaLocation":448},"/pricing/premium/","why premium",{"text":477,"config":478},"Why Ultimate?",{"href":479,"dataGaName":480,"dataGaLocation":448},"/pricing/ultimate/","why ultimate",[482],{"title":483,"links":484},"Contact Us",[485,488,490,492,497,502,507],{"text":486,"config":487},"Contact sales",{"href":35,"dataGaName":36,"dataGaLocation":448},{"text":341,"config":489},{"href":343,"dataGaName":344,"dataGaLocation":448},{"text":346,"config":491},{"href":348,"dataGaName":349,"dataGaLocation":448},{"text":493,"config":494},"Status",{"href":495,"dataGaName":496,"dataGaLocation":448},"https://status.gitlab.com/","status",{"text":498,"config":499},"Terms of use",{"href":500,"dataGaName":501,"dataGaLocation":448},"/terms/","terms of use",{"text":503,"config":504},"Privacy statement",{"href":505,"dataGaName":506,"dataGaLocation":448},"/privacy/","privacy statement",{"text":508,"config":509},"Cookie preferences",{"dataGaName":510,"dataGaLocation":448,"id":511,"isOneTrustButton":14},"cookie preferences","ot-sdk-btn",{"title":71,"links":513,"subMenu":522},[514,518],{"text":515,"config":516},"DevSecOps platform",{"href":53,"dataGaName":517,"dataGaLocation":448},"devsecops platform",{"text":519,"config":520},"AI-Assisted Development",{"href":60,"dataGaName":521,"dataGaLocation":448},"ai-assisted development",[523],{"title":524,"links":525},"Topics",[526,531,536,541,546,551,556,561],{"text":527,"config":528},"CICD",{"href":529,"dataGaName":530,"dataGaLocation":448},"/topics/ci-cd/","cicd",{"text":532,"config":533},"GitOps",{"href":534,"dataGaName":535,"dataGaLocation":448},"/topics/gitops/","gitops",{"text":537,"config":538},"DevOps",{"href":539,"dataGaName":540,"dataGaLocation":448},"/topics/devops/","devops",{"text":542,"config":543},"Version Control",{"href":544,"dataGaName":545,"dataGaLocation":448},"/topics/version-control/","version control",{"text":547,"config":548},"DevSecOps",{"href":549,"dataGaName":550,"dataGaLocation":448},"/topics/devsecops/","devsecops",{"text":552,"config":553},"Cloud Native",{"href":554,"dataGaName":555,"dataGaLocation":448},"/topics/cloud-native/","cloud native",{"text":557,"config":558},"AI for Coding",{"href":559,"dataGaName":560,"dataGaLocation":448},"/topics/devops/ai-for-coding/","ai for coding",{"text":562,"config":563},"Agentic AI",{"href":564,"dataGaName":565,"dataGaLocation":448},"/topics/agentic-ai/","agentic ai",{"title":567,"links":568},"Solutions",[569,571,573,578,582,585,589,592,594,597,600,605],{"text":112,"config":570},{"href":107,"dataGaName":112,"dataGaLocation":448},{"text":102,"config":572},{"href":85,"dataGaName":86,"dataGaLocation":448},{"text":574,"config":575},"Agile development",{"href":576,"dataGaName":577,"dataGaLocation":448},"/solutions/agile-delivery/","agile delivery",{"text":579,"config":580},"SCM",{"href":98,"dataGaName":581,"dataGaLocation":448},"source code management",{"text":527,"config":583},{"href":91,"dataGaName":584,"dataGaLocation":448},"continuous integration & delivery",{"text":586,"config":587},"Value stream management",{"href":140,"dataGaName":588,"dataGaLocation":448},"value stream management",{"text":532,"config":590},{"href":591,"dataGaName":535,"dataGaLocation":448},"/solutions/gitops/",{"text":150,"config":593},{"href":152,"dataGaName":153,"dataGaLocation":448},{"text":595,"config":596},"Small business",{"href":157,"dataGaName":158,"dataGaLocation":448},{"text":598,"config":599},"Public sector",{"href":162,"dataGaName":163,"dataGaLocation":448},{"text":601,"config":602},"Education",{"href":603,"dataGaName":604,"dataGaLocation":448},"/solutions/education/","education",{"text":606,"config":607},"Financial services",{"href":608,"dataGaName":609,"dataGaLocation":448},"/solutions/finance/","financial services",{"title":170,"links":611},[612,614,616,618,621,623,625,627,629,631,633,635],{"text":182,"config":613},{"href":184,"dataGaName":185,"dataGaLocation":448},{"text":187,"config":615},{"href":189,"dataGaName":190,"dataGaLocation":448},{"text":192,"config":617},{"href":194,"dataGaName":195,"dataGaLocation":448},{"text":197,"config":619},{"href":199,"dataGaName":620,"dataGaLocation":448},"docs",{"text":220,"config":622},{"href":222,"dataGaName":223,"dataGaLocation":448},{"text":215,"config":624},{"href":217,"dataGaName":218,"dataGaLocation":448},{"text":230,"config":626},{"href":232,"dataGaName":233,"dataGaLocation":448},{"text":238,"config":628},{"href":240,"dataGaName":241,"dataGaLocation":448},{"text":243,"config":630},{"href":245,"dataGaName":246,"dataGaLocation":448},{"text":248,"config":632},{"href":250,"dataGaName":251,"dataGaLocation":448},{"text":253,"config":634},{"href":255,"dataGaName":256,"dataGaLocation":448},{"text":258,"config":636},{"href":260,"dataGaName":261,"dataGaLocation":448},{"title":272,"links":638},[639,641,643,645,647,649,651,655,660,662,664,666],{"text":279,"config":640},{"href":281,"dataGaName":274,"dataGaLocation":448},{"text":284,"config":642},{"href":286,"dataGaName":287,"dataGaLocation":448},{"text":292,"config":644},{"href":294,"dataGaName":295,"dataGaLocation":448},{"text":297,"config":646},{"href":299,"dataGaName":300,"dataGaLocation":448},{"text":302,"config":648},{"href":304,"dataGaName":305,"dataGaLocation":448},{"text":307,"config":650},{"href":309,"dataGaName":310,"dataGaLocation":448},{"text":652,"config":653},"Sustainability",{"href":654,"dataGaName":652,"dataGaLocation":448},"/sustainability/",{"text":656,"config":657},"Diversity, inclusion and belonging (DIB)",{"href":658,"dataGaName":659,"dataGaLocation":448},"/diversity-inclusion-belonging/","Diversity, inclusion and belonging",{"text":312,"config":661},{"href":314,"dataGaName":315,"dataGaLocation":448},{"text":322,"config":663},{"href":324,"dataGaName":325,"dataGaLocation":448},{"text":327,"config":665},{"href":329,"dataGaName":330,"dataGaLocation":448},{"text":667,"config":668},"Modern Slavery Transparency Statement",{"href":669,"dataGaName":670,"dataGaLocation":448},"https://handbook.gitlab.com/handbook/legal/modern-slavery-act-transparency-statement/","modern slavery transparency statement",{"items":672},[673,676,679],{"text":674,"config":675},"Terms",{"href":500,"dataGaName":501,"dataGaLocation":448},{"text":677,"config":678},"Cookies",{"dataGaName":510,"dataGaLocation":448,"id":511,"isOneTrustButton":14},{"text":680,"config":681},"Privacy",{"href":505,"dataGaName":506,"dataGaLocation":448},241,{"id":684,"title":685,"authorSlugs":686,"authors":688,"body":690,"category":9,"categorySlug":9,"config":691,"content":694,"date":697,"description":695,"extension":12,"externalUrl":6,"featured":14,"heroImage":700,"isFeatured":14,"meta":701,"navigation":14,"path":702,"publishedDate":697,"rawbody":703,"seo":704,"slug":693,"stem":706,"tagSlugs":707,"tags":710,"template":692,"updatedDate":6,"__hash__":711},"blogPosts/en-us/blog/prepare-your-pipeline-for-ai-discovered-zero-days.yml","Prepare your pipeline for AI-discovered zero-days",[687],"omer-azaria",[689],"Omer Azaria","Anthropic's [Mythos Preview model](https://red.anthropic.com/2026/mythos-preview/) recently identified thousands of zero-day vulnerabilities across every major operating system and web browser, including an OpenBSD bug that went undetected for 27 years. In testing, Mythos autonomously chained four vulnerabilities into a working browser exploit that escaped its sandbox. Anthropic is restricting access to Mythos, but the company’s head of offensive cyber research expects threats to have comparable tooling within six to twelve months.\n\nThe defender side of the equation hasn't kept pace. One third of exploited Common Vulnerabilities and Exposures (CVEs) in the first half of 2025 showed activity on or before disclosure day, before most teams even know there's something to patch. AI is compressing that window further, accelerating attackers and flooding teams with whitehat disclosures faster than they can triage. Defender tooling has improved, but most organizations can't operationalize it fast enough to close the gap between discovery and exploitation.\n\nWhen the window between disclosure and exploitation is measured in hours, the security team can't be the last line of defense. Security has to run where code enters the system: in the pipeline, on every merge request, enforced by policy. The fixes that can be automated should be. The ones that can't need to reach the right human faster than they do today.\n\n## Known vulnerabilities are already outpacing remediation\n\nThe bottleneck isn't detection, it's acting at scale on what teams already know. Sixty percent of breaches in the 2025 Verizon DBIR involved exploiting known vulnerabilities where a patch was already available. Teams couldn’t close them in time.\n\nThe backlog was untenable before Mythos. Developers spend [11 hours per month remediating vulnerabilities](https://about.gitlab.com/resources/developer-survey/) post-release instead of shipping new work. Over half of organizations have at least one open internet-facing vulnerability, and the median time to close half of those is 361 days. Exploitation takes hours, while remediation takes months.\n\nAI-assisted development is widening the gap, and stakeholders know it. By June 2025, AI-generated code was adding over 10,000 new security findings per month across Fortune 50 repositories, a 10x jump from six months earlier. Georgia Tech identified 34 [CVEs attributable to AI-generated code](https://research.gatech.edu/bad-vibes-ai-generated-code-vulnerable-researchers-warn) in March 2026, up from 6 in January, and that count reflects only the ones where AI authorship is clear. AI coding assistants hallucinate package names, reach for outdated patterns, and copy insecure examples from training data. More code, more dependencies, and more vulnerabilities per line are generated faster than security teams can review them.\n\nDefenders need to harness frontier AI models, too — not bolted onto the SDLC as external tooling, but running inside the same policies, approvals, and audit trail as the rest of the team. \n\n## Security at the speed of AI coding\n\nWhen a critical CVE drops, how quickly can your team confirm which projects are affected? How many tools does an alert cross before a developer can submit a fix?\n\nThe teams that benefit most from AI already have policies, enforcement, and controls embedded in their development workflows. AI amplifies that foundation. It doesn't replace it.\n\n**Enforcement at the point of change.** As exploitation windows compress, every line of code entering a repository needs to pass through a defined set of controls. Not a separate review, in a different tool, by a different team. Organizations need the ability to enforce security policies across every group and project, with the merge request as the enforcement point. Policies defined once, applied everywhere, with exceptions reviewed, approved, and logged.\n\n**Simple issues caught before the merge request, not during.** Hardcoded secrets, known-vulnerable imports, and deprecated API calls can be flagged in the IDE before a developer pushes a commit. Catching them at authoring time means fewer findings blocking the MR, so review cycles go to the findings that require cross-component context: reachability, exploitability, and architectural risk.\n\n**Triage automated by default, not by exception.** Embedding security into every merge request creates a volume problem. More scans, more findings, more noise reaching developers who aren’t trained to distinguish a reachable critical from a theoretical one. AI must handle false positive detection, reachability, exploitability context, and severity assessment before a developer sees the finding, so the findings they see actually warrant their time.\n\n**Remediation governed like any other change.** AI-based remediation compresses the timeline for closing vulnerabilities, but every generated fix must move through the same governance as a human-authored change: policies enforce scans, the right reviewers approve, and evidence is recorded. GitLab’s automated remediation capability proposes each fix in a merge request with a confidence score. The MR records which policy applied, which scans ran, what they found, and who approved. Human code and AI-generated code move through the same process, with the same audit trail.\n\n## What a ready pipeline looks like\n\nHere's how these pieces work together when a high-severity vulnerability is discovered and the clock is running.\n\nA proof-of-concept exploit for a vulnerability in a popular open-source package appears on a security mailing list. There’s no CVE, no National Vulnerability Database (NVD) entry, and no scanner signature yet. The security team finds out the usual way: someone shares it in Slack.\n\nA security engineer asks the security agent if the package is in use, which projects have affected versions, and whether any vulnerable call paths are reachable in production. The agent checks the dependency graph for every project, matches the affected versions and entry points from the disclosure, and returns a ranked list of exposed projects with details about reachability. There’s no need to search through repositories by hand or wait for a scanner update. The question, \"Are we exposed?\" is answered in minutes.\n\nThe engineer starts a remediation campaign for every exposed project. The remediation agent suggests fixes: version updates where a patched release is available, and targeted call-path patches where it is not. Scan execution policies are already in place for projects tagged SOC 2. The engineer hardens the rules to block merges on any merge request that introduces or keeps the affected dependency, and an approval policy now requires security sign-off on every fix. The agent's first proposed patch fails the pipeline when an integration test catches a regression. The agent revises the patch based on the test failure, and the second attempt passes. Developers review the changes, security signs off under the stricter policy, and merges proceed across the campaign.\n\nAt the next audit review, the security team presents a report showing how policies were enforced and risks were reduced during the campaign. It includes scan results, policies applied, approvers, and merge timestamps for every MR in every affected project. The evidence was automatically generated in flight, not assembled after the fact.\n\n## Close the gaps now\n\nMythos exists today, and comparable models will be in attacker hands within a year. Every month between now and then is a chance to strengthen your software supply chain.\n\nAsk these questions about your pipeline:\n\n* How do you enforce that security scans run on every merge request, not just the projects where teams configured them?\n\n* If a compromised package entered your dependency tree today, would your pipeline catch it before build?\n\n* When a scanner flags a critical finding, how many tool boundaries does it cross before a developer starts the fix?\n\n* If an AI agent proposed a code fix for a vulnerability, what process would that fix go through before reaching production, and is that process auditable?\n\n* When auditors ask for evidence that a specific policy was enforced on a specific change, how long does it take to produce?\n\nIf the answers expose gaps, address them now. [Talk to a GitLab solutions architect](https://about.gitlab.com/sales/) about the role of security governance in your development lifecycle.",{"featured":14,"template":692,"slug":693},"BlogPost","prepare-your-pipeline-for-ai-discovered-zero-days",{"title":685,"description":695,"authors":696,"date":697,"body":690,"category":9,"tags":698,"heroImage":700},"AI is finding vulnerabilities faster than teams can patch. Learn how pipeline enforcement, automated triage, and AI remediation close the gap.",[689],"2026-04-20",[699,9,515],"AI/ML","https://res.cloudinary.com/about-gitlab-com/image/upload/v1772195014/ooezwusxjl1f7ijfmbvj.png",{},"/en-us/blog/prepare-your-pipeline-for-ai-discovered-zero-days","seo:\n  config:\n    noIndex: false\n  title: Prepare your pipeline for AI-discovered zero-days\n  description: AI is finding vulnerabilities faster than teams can patch. Learn\n    how pipeline enforcement, automated triage, and AI remediation close the\n    gap.\ncontent:\n  title: Prepare your pipeline for AI-discovered zero-days\n  description: AI is finding vulnerabilities faster than teams can patch. Learn\n    how pipeline enforcement, automated triage, and AI remediation close the\n    gap.\n  authors:\n    - Omer Azaria\n  date: 2026-04-20\n  body: >-\n    Anthropic's [Mythos Preview\n    model](https://red.anthropic.com/2026/mythos-preview/) recently identified\n    thousands of zero-day vulnerabilities across every major operating system\n    and web browser, including an OpenBSD bug that went undetected for 27 years.\n    In testing, Mythos autonomously chained four vulnerabilities into a working\n    browser exploit that escaped its sandbox. Anthropic is restricting access to\n    Mythos, but the company’s head of offensive cyber research expects threats\n    to have comparable tooling within six to twelve months.\n\n\n    The defender side of the equation hasn't kept pace. One third of exploited Common Vulnerabilities and Exposures (CVEs) in the first half of 2025 showed activity on or before disclosure day, before most teams even know there's something to patch. AI is compressing that window further, accelerating attackers and flooding teams with whitehat disclosures faster than they can triage. Defender tooling has improved, but most organizations can't operationalize it fast enough to close the gap between discovery and exploitation.\n\n\n    When the window between disclosure and exploitation is measured in hours, the security team can't be the last line of defense. Security has to run where code enters the system: in the pipeline, on every merge request, enforced by policy. The fixes that can be automated should be. The ones that can't need to reach the right human faster than they do today.\n\n\n    ## Known vulnerabilities are already outpacing remediation\n\n\n    The bottleneck isn't detection, it's acting at scale on what teams already know. Sixty percent of breaches in the 2025 Verizon DBIR involved exploiting known vulnerabilities where a patch was already available. Teams couldn’t close them in time.\n\n\n    The backlog was untenable before Mythos. Developers spend [11 hours per month remediating vulnerabilities](https://about.gitlab.com/resources/developer-survey/) post-release instead of shipping new work. Over half of organizations have at least one open internet-facing vulnerability, and the median time to close half of those is 361 days. Exploitation takes hours, while remediation takes months.\n\n\n    AI-assisted development is widening the gap, and stakeholders know it. By June 2025, AI-generated code was adding over 10,000 new security findings per month across Fortune 50 repositories, a 10x jump from six months earlier. Georgia Tech identified 34 [CVEs attributable to AI-generated code](https://research.gatech.edu/bad-vibes-ai-generated-code-vulnerable-researchers-warn) in March 2026, up from 6 in January, and that count reflects only the ones where AI authorship is clear. AI coding assistants hallucinate package names, reach for outdated patterns, and copy insecure examples from training data. More code, more dependencies, and more vulnerabilities per line are generated faster than security teams can review them.\n\n\n    Defenders need to harness frontier AI models, too — not bolted onto the SDLC as external tooling, but running inside the same policies, approvals, and audit trail as the rest of the team. \n\n\n    ## Security at the speed of AI coding\n\n\n    When a critical CVE drops, how quickly can your team confirm which projects are affected? How many tools does an alert cross before a developer can submit a fix?\n\n\n    The teams that benefit most from AI already have policies, enforcement, and controls embedded in their development workflows. AI amplifies that foundation. It doesn't replace it.\n\n\n    **Enforcement at the point of change.** As exploitation windows compress, every line of code entering a repository needs to pass through a defined set of controls. Not a separate review, in a different tool, by a different team. Organizations need the ability to enforce security policies across every group and project, with the merge request as the enforcement point. Policies defined once, applied everywhere, with exceptions reviewed, approved, and logged.\n\n\n    **Simple issues caught before the merge request, not during.** Hardcoded secrets, known-vulnerable imports, and deprecated API calls can be flagged in the IDE before a developer pushes a commit. Catching them at authoring time means fewer findings blocking the MR, so review cycles go to the findings that require cross-component context: reachability, exploitability, and architectural risk.\n\n\n    **Triage automated by default, not by exception.** Embedding security into every merge request creates a volume problem. More scans, more findings, more noise reaching developers who aren’t trained to distinguish a reachable critical from a theoretical one. AI must handle false positive detection, reachability, exploitability context, and severity assessment before a developer sees the finding, so the findings they see actually warrant their time.\n\n\n    **Remediation governed like any other change.** AI-based remediation compresses the timeline for closing vulnerabilities, but every generated fix must move through the same governance as a human-authored change: policies enforce scans, the right reviewers approve, and evidence is recorded. GitLab’s automated remediation capability proposes each fix in a merge request with a confidence score. The MR records which policy applied, which scans ran, what they found, and who approved. Human code and AI-generated code move through the same process, with the same audit trail.\n\n\n    ## What a ready pipeline looks like\n\n\n    Here's how these pieces work together when a high-severity vulnerability is discovered and the clock is running.\n\n\n    A proof-of-concept exploit for a vulnerability in a popular open-source package appears on a security mailing list. There’s no CVE, no National Vulnerability Database (NVD) entry, and no scanner signature yet. The security team finds out the usual way: someone shares it in Slack.\n\n\n    A security engineer asks the security agent if the package is in use, which projects have affected versions, and whether any vulnerable call paths are reachable in production. The agent checks the dependency graph for every project, matches the affected versions and entry points from the disclosure, and returns a ranked list of exposed projects with details about reachability. There’s no need to search through repositories by hand or wait for a scanner update. The question, \"Are we exposed?\" is answered in minutes.\n\n\n    The engineer starts a remediation campaign for every exposed project. The remediation agent suggests fixes: version updates where a patched release is available, and targeted call-path patches where it is not. Scan execution policies are already in place for projects tagged SOC 2. The engineer hardens the rules to block merges on any merge request that introduces or keeps the affected dependency, and an approval policy now requires security sign-off on every fix. The agent's first proposed patch fails the pipeline when an integration test catches a regression. The agent revises the patch based on the test failure, and the second attempt passes. Developers review the changes, security signs off under the stricter policy, and merges proceed across the campaign.\n\n\n    At the next audit review, the security team presents a report showing how policies were enforced and risks were reduced during the campaign. It includes scan results, policies applied, approvers, and merge timestamps for every MR in every affected project. The evidence was automatically generated in flight, not assembled after the fact.\n\n\n    ## Close the gaps now\n\n\n    Mythos exists today, and comparable models will be in attacker hands within a year. Every month between now and then is a chance to strengthen your software supply chain.\n\n\n    Ask these questions about your pipeline:\n\n\n    * How do you enforce that security scans run on every merge request, not just the projects where teams configured them?\n\n\n    * If a compromised package entered your dependency tree today, would your pipeline catch it before build?\n\n\n    * When a scanner flags a critical finding, how many tool boundaries does it cross before a developer starts the fix?\n\n\n    * If an AI agent proposed a code fix for a vulnerability, what process would that fix go through before reaching production, and is that process auditable?\n\n\n    * When auditors ask for evidence that a specific policy was enforced on a specific change, how long does it take to produce?\n\n\n    If the answers expose gaps, address them now. [Talk to a GitLab solutions architect](https://about.gitlab.com/sales/) about the role of security governance in your development lifecycle.\n  category: security\n  tags:\n    - AI/ML\n    - security\n    - DevSecOps platform\n  heroImage: https://res.cloudinary.com/about-gitlab-com/image/upload/v1772195014/ooezwusxjl1f7ijfmbvj.png\nconfig:\n  featured: true\n  template: BlogPost\n  slug: prepare-your-pipeline-for-ai-discovered-zero-days\n",{"config":705,"title":685,"description":695},{"noIndex":10},"en-us/blog/prepare-your-pipeline-for-ai-discovered-zero-days",[708,9,709],"aiml","devsecops-platform",[699,9,515],"O9D6KixG6qt3e8V6INxYHXq6exvU9DpD1iBjYlFV9CY",[713,722,731,740,749,758,767,776,783],{"content":714,"config":720},{"title":715,"heroImage":716,"category":9,"description":717,"authors":718},"Manage vulnerability noise at scale with auto-dismiss policies","https://res.cloudinary.com/about-gitlab-com/image/upload/v1774375772/kpaaaiqhokevxxeoxvu0.png","Learn how to cut through scanner noise and focus on the vulnerabilities that matter most with GitLab security, including use cases and templates.",[719],"Grant Hickman",{"externalUrl":-1,"slug":721},"auto-dismiss-vulnerability-management-policy",{"content":723,"config":729},{"title":724,"heroImage":725,"category":9,"description":726,"authors":727},"GitLab 18.10 brings AI-native triage and remediation ","https://res.cloudinary.com/about-gitlab-com/image/upload/v1773843921/rm35fx4gylrsu9alf2fx.png","Learn about GitLab Duo Agent Platform capabilities that cut noise, surface real vulnerabilities, and turn findings into proposed fixes.",[728],"Alisa Ho",{"externalUrl":-1,"slug":730},"gitlab-18-10-brings-ai-native-triage-and-remediation",{"content":732,"config":738},{"title":733,"heroImage":734,"category":9,"description":735,"authors":736},"A complete guide to GitLab Container Scanning","https://res.cloudinary.com/about-gitlab-com/image/upload/v1772721753/frfsm1qfscwrmsyzj1qn.png","Explore GitLab's various container scanning methods and learn how to secure containers at every lifecycle stage.",[737],"Fernando Diaz",{"externalUrl":-1,"slug":739},"complete-guide-to-gitlab-container-scanning",{"content":741,"config":747},{"title":742,"heroImage":743,"category":9,"description":744,"authors":745},"Track vulnerability remediation with the updated GitLab Security Dashboard","https://res.cloudinary.com/about-gitlab-com/image/upload/v1771438388/t6sts5qw4z8561gtlxiq.png","Quickly prioritize remediation on high-risk projects and measure progress with vulnerability insights.",[728,746],"Mike Clausen",{"externalUrl":-1,"slug":748},"track-vulnerability-remediation-with-the-updated-gitlab-security-dashboard",{"content":750,"config":756},{"title":751,"heroImage":752,"category":9,"description":753,"authors":754},"How to set up GitLab SAML SSO with Google Workspace","https://res.cloudinary.com/about-gitlab-com/image/upload/v1759320418/xjmqcozxzt4frx0hori3.png","Learn how to automate user provisioning and sync permissions with Google groups with this step-by-step guide.",[755],"Omid Khan",{"externalUrl":-1,"slug":757},"how-to-set-up-gitlab-saml-sso-with-google-workspace",{"content":759,"config":765},{"title":760,"heroImage":761,"category":9,"description":762,"authors":763},"GitLab Bug Bounty Program policy updates","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749664844/Blog/Hero%20Images/AdobeStock_941867776.jpg","Learn about the program's enhanced clarity and updated scope.",[764],"Kayla Hagopian",{"externalUrl":-1,"slug":766},"gitlab-bug-bounty-program-policy-updates",{"content":768,"config":774},{"title":769,"heroImage":770,"category":9,"description":771,"authors":772},"Strengthening GitLab.com security: Mandatory multi-factor authentication","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749664923/Blog/Hero%20Images/security-checklist.png","Learn how GitLab is implementing mandatory MFA as part of Secure by Design commitment and what it means for users.",[773],"Kim Waters",{"externalUrl":-1,"slug":775},"strengthening-gitlab-com-security-mandatory-multi-factor-authentication",{"content":777,"config":781},{"title":778,"heroImage":752,"category":9,"description":779,"authors":780},"OWASP Top 10 2025: What's changed and why it matters","Explore new supply chain and error handling risks, ranking shifts, and remediation strategies for all 10 categories.",[737],{"externalUrl":-1,"slug":782},"2025-owasp-top-10-whats-changed-and-why-it-matters",{"content":784,"config":789},{"title":785,"heroImage":786,"category":9,"description":787,"authors":788},"AI-powered vulnerability triaging with GitLab Duo Security Agent","https://res.cloudinary.com/about-gitlab-com/image/upload/v1756122536/akivvcnafog9c4dhhzkp.png","Learn how this GitLab Duo Agent Platform capability uses AI to prioritize vulnerabilities, reduce alert fatigue, and help teams focus on critical security risks.",[737],{"externalUrl":-1,"slug":790},"vulnerability-triage-made-simple-with-gitlab-security-analyst-agent",1777493617962]