A new year, a smarter platform, and what’s next for intelligent software delivery
January 2026 Monday Merge
Hi GitLab community. It’s Fatima here, and welcome to the first Monday Merge of 2026, hot off the processor!
A new year always brings fresh momentum, and this one is starting strong. We’re kicking things off with the GitLab 18.7 release, then a customer story that shows what real DevSecOps velocity looks like at scale, an open source hackathon you can jump into this month, and a couple of thought-provoking reads on what 2026 has in store for developer teams and AI-driven organizations.
Let’s gooooooo 🚀🚀🚀🚀🚀
GitLab 18.7 is here
GitLab 18.7 continues to advance our vision for intelligent, orchestrated software delivery with practical development, operations, and security capabilities that strengthen control, improve consistency, and build confidence as teams integrate AI into their workflows.
This release adds important building blocks as we move toward the general availability of the GitLab Duo Agent Platform coming in 18.8. You’ll see new automation capabilities that help teams define repeatable, multi-step workflows using Custom Flows, allowing agents to diagnose failed pipelines, update dependencies, or run policy checks automatically when key events occur.
Security teams will appreciate AI-powered SAST false positive detection, which helps surface real risks faster and reduces time spent triaging noise. And with Custom Agent Versioning, teams can now pin agents and flows to specific versions, bringing much needed stability and control to AI-powered workflows.
We’ve also introduced a new Data Analyst Agent, making it easier to explore GitLab data using natural language and reusable queries, along with granular settings that give administrators control over which foundational agents are available across their organization.
On the CI/CD side, dynamic input selection simplifies pipeline triggers through intuitive dropdowns, while new guardrails for CI/CD Catalog publishing help enterprise teams maintain a trusted ecosystem of reusable components. Policy guidance also gets more flexible with Warn Mode for merge request approval policies, allowing teams to introduce or refine rules without blocking delivery.
And if you want to see where all of this is heading next, join us live on February 10, 2026, for a global launch event that brings our vision of GitLab as the intelligent orchestration platform to life. You’ll see intelligent DevSecOps workflows in action and hear how teams are tackling the AI paradox - where AI makes coding faster but can slow down the rest of the software lifecycle - while staying in flow.
Key highlights in 18.7 include:
- Custom Flows for agent orchestration: Define YAML-based, multi-step workflows that coordinate agents to diagnose failures, fix issues, apply policy checks, and notify teams automatically.
- Pinned versions for custom agents and flows: Lock agents to specific versions to prevent breaking changes and support safer testing, staged upgrades, and production stability.
- AI-assisted SAST false positive detection: Reduce security noise by identifying likely false positives earlier, keeping vulnerability status consistent across pipelines and merge requests.
- Data Analyst Agent for GitLab insights: Ask natural language questions about your GitLab data and get instant answers, generated queries, and reusable insights without dashboards or manual query writing.
- Warn Mode for merge request approval policies: Introduce or refine policies without blocking merges, helping teams assess impact before enforcing rules at scale.
Customer spotlight: Ericsson
This month’s customer spotlight highlights Ericsson, where GitLab is helping deliver faster, more reliable deployments of mission-critical OSS and BSS systems used by communications service providers around the world.
By standardizing on GitLab Premium and enabling GitOps-based delivery, Ericsson cut deployment times by 50 percent, saved over 130,000 hours in just six months, and increased testing coverage tenfold. What once took months now happens in weeks, without sacrificing the reliability required for systems where downtime directly impacts revenue.
As Daniel Costa Soares, Head of Software Automation and Support for Ericsson OSS/BSS, puts it, the faster deployment capabilities GitLab enables directly contribute to business growth by allowing teams to deliver innovation at the speed their markets demand.
It’s a powerful example of how consolidating tooling, increasing automation, and gaining end-to-end visibility can transform both developer experience and customer outcomes.
Recommended by LinkedIn
Upcoming event: GitLab Hackathon
If you’re looking for a hands-on way to start the year, the GitLab Hackathon runs from January 22 to January 29.
This week-long virtual event is open to everyone and focused on collaboration, learning, and contribution across the GitLab ecosystem. Whether you’re coding, improving documentation, contributing translations, or jumping into UX and design work, every merged merge request counts.
You’ll earn points, unlock profile achievements, and collaborate with a global open source community. If you’ve ever wanted to contribute to GitLab, this is a great place to start.🔗 Sign up and contribute!
What we’re reading
We’re starting 2026 with a couple of reads that look ahead at what developer teams and AI-driven organizations should prepare for next.
In The Year AI Moves from Pilots to Practice: Three Predictions for Government in 2026, Bob Stevens explores how government agencies are moving from AI pilots to real implementation. He highlights how AI-powered modernization can turn multi-year legacy system transformations into focused, months-long sprints, while shifting security from reactive to proactive and expanding who can build software across organizations.
Meanwhile, Michelle Gill’s piece The Hidden Truths About Building and Retaining GenAI Teams digs into the human side of AI innovation. One insight that stands out is that hiring great AI talent is only the beginning. Sustained success comes from clear decision making frameworks, fast execution, and giving teams problems worth solving so they can ship at the pace this space demands.
Both are worth your time if you’re thinking about how to scale AI responsibly and effectively in 2026.
A quote to start the year
We’ll close this first Monday Merge of the year with a reminder that feels especially relevant as teams take on new challenges and opportunities.
“Unity is strength. When there is teamwork and collaboration, wonderful things can be achieved.” Mattie Stepanek
Here’s to building together, learning together, and shipping great things in the year ahead.
Thanks for reading, and as always, happy merging!
Fatima Sarah Khalid | Developer Advocate, GitLab
Like this newsletter? Forward it to your team. And don’t forget to 👉 subscribe to our YouTube channel!
😃
The “new Data Analyst Agent” is interesting to me. Has anyone had large scale/real world success with this yet? This could help when a codebase reaches a critical level of complexity and needs major refactors across unrelated areas, but AI analysis isn’t integrated at scale due to time and budget constraints. This is especially common in microservice architectures, where changes can span several services across different teams and are hard to assess quickly without pulling everything locally and using IDEs and tooling to scope effort duration and communicate impact to non-technical teams.
Great. Good direction. Looking forward to the next.
Good perspective. What stood out to me is the shift from optimizing individual tools to optimizing the delivery system as a whole. The real leverage seems to be in feedback loops, how fast signals from prod reach planning and execution. Intelligence there beats any single smart component.
Wow🚀