How Gitlab manages AI’s emissions overhead

Gitlab, which sells software coding tools used by more than half of Fortune 500 companies including Nvidia and Goldman Sachs, is adding artificial intelligence features across its product portfolio and using them to improve productivity.

Environmental considerations were integrated into both plans, and Gitlab adopted guidelines in 2025 that define what AI software vendors should disclose as part of sales contracts. Gitlab’s chief information officer helped write those rules.

“We don’t want to discourage AI use,” said Stacy Cline, senior director of sustainability at Gitlab. “Our goal is to encourage intentional use, since efficient prompts and workflows reduce both cost and energy consumption. It’s important that employees understand the energy and resources behind the tools they’re using every day.”

Gitlab’s “Green DevOps” policy, which incorporates sustainability considerations into software app design processes, considers factors such as the carbon footprint associated with different cloud computing services used for training or to run queries.

“Within sustainable AI, we focus on understanding its carbon cost, encouraging responsible use, getting the right data and systems set up to measure our emissions and being transparent about our work,” Cline said.

Gitlab’s direct emissions are negligible. Its top sources of greenhouse gas emissions are purchased goods and services, especially cloud software from Amazon Web Services and Google, and business travel.

These categories account for 52 percent and 43 percent of Gitlab’s overall carbon footprint, respectively, according to its 2025 sustainability report. (Based on past publishing dates, Gitlab’s next report is due in mid-July.)  

Gitlab pledged to get 70 percent of its biggest-emitting suppliers to set science-based emissions reduction targets by 2029. It hasn’t provided a recent progress report about that goal, but is talking about next steps with more than 85 of its closest partners.

AI-boosted shortcuts

Gitlab’s sustainability team created several AI agents to support its agenda, including a tool to review “hundreds” of requests for proposals and customer questionnaires.

To help tame the company’s large travel-related footprint, the team compiled a list of event sustainability suggestions and wrote an AI prompt that can be used by employees planning events to compare the carbon impact of various destinations. The information is part of Gitlab’s corporate handbook.

The team also used NotebookLM, which uses specific documents to generate responses, to digest information from lengthy sustainability and ESG disclosures and create specialized content for its sales team. The output was a mini-podcast that suggests five talking points that a salesperson can use with customers that have corporate sustainability commitments.

“By doing this, we ensured that the most relevant content reaches the right audiences in a format that’s easy to digest,” Cline said

Gitlab is also encouraging other developers — its tools have more than 50 million registered users — to use AI as the foundation for new corporate sustainability applications. 

The company hosts quarterly hackathons, which now include a “green agent” category. The winning green agent for the hackathon hosted in February and March of this year was Green Pipe, which estimates carbon emissions associated with design choices made during software development processes.