Updated

Updated · OpenAI · Oct 26

CodeRabbit Deploys GPT-6.1 Sol to Scan Thousands of Repositories for Security Risks

Updated

Updated · OpenAI · Oct 26

CodeRabbit Deploys GPT-6.1 Sol to Scan Thousands of Repositories for Security Risks

1 articles · Updated · OpenAI · Oct 26

CodeRabbit has made GPT-6.1 Sol the primary model in its security agent, using it to detect security issues across thousands of repositories and organizations.
GPT-6.1 Sol extends review beyond a single pull request by tracing how code changes affect shared libraries, dependencies and safeguards across files and components.
In Deep Scan and CodeRabbit Secure, the model investigates whole repositories and linked codebases—such as shared authentication libraries—to map how a weakness could spread through a wider system.
For selected findings, CodeRabbit Secure can verify potential exploits in an isolated sandbox and, for supported issues, draft a pull request with a proposed fix for human approval.
The rollout reflects a broader push to pair frontier AI reasoning with existing scanners while keeping developers responsible for deciding which changes reach production.

Sources

If an AI acts as a security guard, who prevents the AI itself from introducing subtle new vulnerabilities into the codebase?

Will automated AI pull requests actually reduce developer workload, or merely shift the burden to reviewing complex machine-generated code?

Can an isolated sandbox truly replicate complex production environments to prove a vulnerability's exploitability without risking actual data?

1 total