Application security where found means fixed
Find vulnerabilities in any language with CodeQL and AI Scan, then hand the alerts to Copilot, one at a time or straight from your backlog.
70% Reduction in mean time to remediate
100K+ Vulnerabilities fixed in six months
1.5M Developer hours saved
Detect and remediate vulnerabilities early with AI-powered fixes
CodeQL catches known patterns fast and at low cost, with the same answer every run. AI Scan goes where rules can't reach, including languages CodeQL doesn't cover and risks unique to AI applications.
Learn more about CodeQL
Assign one alert or up to 25 to Copilot. It explores the codebase, makes the change, and opens a draft pull request for a developer to review.
Learn about agentic autofix
Scan the code that merged before anyone checked it, then work down the backlog with security campaigns and agentic autofix.
Apply security settings to every repository with one configuration, and use rulesets to stop merges that add high-severity alerts. Track coverage in security overview.
Learn how to enable security at scale
Evaluate exposure to application vulnerabilities and leaked secrets in your codebase with our free risk assessment tool.
Explore GitHub security risk assessments
Build secure software from day one
Security should be built in, not bolted on. With Code Security, you can find, fix, and prevent vulnerabilities seamlessly—keeping your software resilient from development to deployment.
Best practices for more secure software
Discover developer-first security
Take an in-depth look at the current state of application security.
Explore the DevSecOps guide
Learn how to write more secure code from the start with DevSecOps.
Avoid AppSec pitfalls
Explore common application security pitfalls and how to avoid them.
Frequently Asked Questions
What is Code Security?
GitHub Code Security helps developers secure their code without slowing down. CodeQL and AI Scan find vulnerabilities, Copilot fixes them, and dependency review and security campaigns keep risk from piling up, all inside the GitHub workflow.
What is Copilot Autofix?
Copilot Autofix uses AI-powered code suggestions to automatically fix security vulnerabilities identified by CodeQL. When a security vulnerability is detected, Copilot Autofix analyzes the code context, understands the underlying security issue, and generates a precise, contextually appropriate fix. This feature bridges the gap between vulnerability detection and remediation, enabling developers to review and apply AI-suggested fixes directly within their workflow.
What are Security Campaigns?
Security campaigns provide a structured framework for planning, tracking, and implementing security fixes across multiple repositories and teams allowing you to systematically burn down security debt. With With security campaigns, security teams can group related vulnerabilities, prioritize remediation efforts, assign ownership, and monitor progress through a unified dashboard. Security campaigns can be organized by vulnerability type, security initiative, compliance requirement, or any other logical grouping to coordinate security improvements at scale.
What is dependency analysis?
Dependency review scans pull requests for vulnerable dependencies before they're introduced into your codebase. It evaluates the security impact of dependency changes, identifying vulnerable packages and their severity levels to prevent security issues from being merged. The tool shows detailed dependency changes by comparing the base and head branches, highlighting added, removed, and updated dependencies along with their known vulnerabilities
What is EPSS?
Dependabot alerts now feature the Exploit Prediction Scoring System (EPSS) from the global Forum of Incident Response and Security Teams (FIRST), helping better assess vulnerability risks. EPSS helps organizations prioritize vulnerability remediation by predicting the likelihood of a vulnerability being exploited in the next 30 days. It provides a score ranging from 0 to 1 (0-100%), alongside a percentile ranking to indicate how the vulnerability compares to others.
What is the code security risk assessment?
The Code Security Risk Assessment is a free evaluation that analyzes repositories to identify potential code-level vulnerabilities and highlight areas where GitHub Code Security can help improve security posture. Learn more about GitHub security risk assessments.
What is AI Scan?
AI Scan uses AI to find vulnerabilities, including in languages CodeQL doesn't support. It reviews every pull request (generally available) and scans whole repositories on demand (public preview). Repository findings become code scanning alerts that you track, assign, and fix like any other. AI Scan is off by default: an admin turns it on in code scanning settings, and anyone with code scanning write access can then start a repository scan.
Can Code Security find vulnerabilities in AI applications?
Yes. Code scanning for AI applications (public preview) reviews pull requests for risks unique to AI apps and agents: indirect prompt injection, sensitive information disclosure, and excessive agency. It flags code paths where untrusted content can reach a model with access to tools or data, and explains how to fix them. Turn it on in code scanning settings.