The latest GitHub Copilot update includes vulnerability filtering

News Summary:

  • The underlying OpenAI Codex AI model has been updated as part of the updates announced on February 14 to support significant enhancements to code quality recommendations. Additionally, the amount of time needed to deliver suggestions has decreased. Both Copilot for Individuals and Copilot for Businesses can use these improvements.

  • Algorithms have been added to GitHub Copilot, the contentious tool that offers developers AI-assisted coding suggestions, to increase the quality and security of those suggestions.

Furthermore, Copilot’s AI-based vulnerability filtering now prevents unsecured coding practices in real time. Hard-coded credentials, path injections, and SQL injections are a few examples of common vulnerable coding patterns that are targeted. According to GitHub, vulnerable patterns can even be found in incomplete fragments.

GitHub also announced this week the general availability of Github Copilot for Business for all businesses, including those that do not use GitHub. This announcement comes after a beta period that began in December. Businesses can sign up for Copilot and assign seats right away. VPN proxy support is also included in GitHub Copilot for Businesses, allowing Copilot to be used in any working environment.

Copilot is an AI pair programming tool that offers auto-complete-style suggestions in a variety of programming languages. It was trained using code from public repositories. Copilot is available through extensions for Visual Studio Code, Visual Studio, Neovim, and the JetBrains IDE suite. users can sign up for a free trial of Copilot.

Copilot has been criticized since its debut in June 2021, with concerns raised about the propriety of using publicly available code to train the AI. These concerns prompted the Free Software Foundation to file a lawsuit and protest. However, GitHub announced this week that more than 400 organizations are already using Copilot, and that the tool is assisting developers in coding faster.