GitLab, Google Cloud Combine for AI-Assisted Platform Tools

DevSecOps platform maker GitLab has announced an extension of its partnership with Google Cloud to bring new — and some experimental — AI-based options to enterprises. The two companies will deploy Google Cloud’s customizable foundation models and open generative AI infrastructure to provide users with AI-assisted features located within the GitLab platform.

Google Cloud users already have generative AI support available on the company’s home-built Vertex AI to build different applications within their own security boundaries. One of GitLab’s goals is to improve developers’ DevSecOps workflow efficiency by 10X, by applying AI-assisted workflows to all users, Google told The New Stack.

“This partnership with Google Cloud enables GitLab to offer private and secure AI-powered features while maintaining customer data in our cloud infrastructure,” David DeSanto, Chief Product Officer at GitLab, said in a media advisory.

New AI-assisted features in the platform

An experimental feature using Google Cloud’s generative AI models is called Explain this Vulnerability. This empowers companies to make security a cross-organizational effort by providing users with a natural language description of vulnerabilities found in their code and a recommendation for how to resolve them at the time of detection, Google said.

Developers, in addition to security and operations teams, can use Explain this Vulnerability, allowing users to stay secure while remaining efficient and improving speed to delivery.

Other new capabilities include:

• Using generative AI support in Vertex AI, GitLab users can tune Google’s foundation models with their own data and use these models to deliver new generative AI-powered experiences.
• Model Garden, offered as part of generative AI support in Vertex AI. This is a single environment to search, discover and interact with Google’s own foundation models — and in time, hundreds of open source and third-party models. In Model Garden, users have access to more than just text models; they will be able to build next-generation applications with access to multimodal models from Google across vision, dialog, code generation, and code completion.
• It is customizable. “In this case, GitLab intends to build upon Google Cloud’s foundational models to customize it to specifically meet the needs of its users,” a Google spokeswoman said. “However, in addition to the foundation models and APIs offered within Vertex AI, Google Cloud also offers generative AI-powered search and conversational experiences with Gen App Builder. This allows developers, even those with limited machine-learning skills, to get started with Google’s foundation models.”

Both companies are utilizing a “privacy-first” approach to deploying AI in these tools, a Google spokeswoman told The New Stack.

“GitLab chose to partner with Google Cloud because of its strong commitment to meet enterprise privacy expectations and its leadership in AI,” she said. “The new AI features enable GitLab to maintain its commitment to protecting user privacy by containing customer intellectual property and source code within GitLab’s cloud infrastructure.”

GitLab’s 2023 DevSecOps Report found that developers are increasingly using AI for testing and security, with 62% of developers using AI/ML to check code, up from 51% in 2022. Additionally, 36% of developers use AI/ML for code review, up from 31% the previous year.

GitLab has made other news recently. On April 24, GitLab released new AI/ML and automated features to integrate security into every stage of its software delivery workflow. These updates are in addition to their generative AI Code Suggestions feature, which is in beta testing now.

With supply chain security attacks like the recent 3CX breach still making waves, security is now at the root of the development lifecycle.