Microsoft adds Amazon capacity for GitHub after AI outages and reliability failures

Microsoft is turning to Amazon for help keeping GitHub running after a spring and summer of AI-driven capacity stress and outages. According to two people familiar with the plans, GitHub is adding extra computing capacity via Amazon Web Services as Microsoft works through a reliability problem caused by an unusually fast surge in AI coding usage.

This is not a small operational tweak for a platform Microsoft acquired in 2018. GitHub used to run its own data centers for the most part, and Microsoft had planned to move the coding platform entirely to its Azure cloud service by 2027. Now the calendar looks less like a straight line and more like a scramble: AI coding demand has swamped GitHub’s compute resources, and that has forced Microsoft to lean on its biggest cloud rival, AWS.

To understand why this matters, you have to look at what GitHub does. It is a popular place for engineers to store and manage code, collaborate on projects, and track changes via commits, which serve as a proxy for development activity. This year, the AI wave changed the shape of that activity. GitHub commits were on pace to reach 14 billion in 2026, up from 1 billion in 2025, Chief Operating Officer Kyle Daigle wrote on X in April. That delta is the whole story in one statistic: AI coding tools make it easier for developers to write more software, and more software means more load on the systems.

When load spikes in this way, the immediate goal is reliability. Microsoft’s spokesperson confirmed GitHub is tapping multiple cloud providers, but declined to comment on any Amazon involvement. The spokesperson also said that an “incredible spike in agentic development that began late last year has tested our infrastructure’s limits.” To meet this demand, Microsoft said it is “both accelerating our move to Azure and continuing to explore a multi-cloud strategy to ensure we have the future capacity, compute elasticity and horizontal scale required to support continued growth.” In plain English: Microsoft wants Azure as the end state, but it needs elastic capacity today, and it is buying that elasticity where it can.

The headline twist is the rivalry. Microsoft is battling AWS for market share in the cloud. So outsourcing capacity to AWS, rather than addressing GitHub needs through Azure, looks counterintuitive for a company that wants to win cloud customers. But the story also includes a broader industry pattern: other big tech companies are striking similar deals because AI demand is overwhelming the usual procurement playbook.

This week’s Amazon help for GitHub reads like a companion to other compute “routing” that companies have announced recently. Earlier this month, SpaceX and Google disclosed a new deal in which Google will pay SpaceX $920 million a month for AI compute capacity from October 2026 to June 2029. The source notes that this emerged just two months after Google’s own cloud business agreed to sell AI compute capacity to Anthropic. Put differently, as AI compute becomes a strategic bottleneck, the competitive map blurs. Even companies that compete aggressively in one market often cooperate or contract across the value chain when the alternative is downtime.

For GitHub, downtime is not a theoretical risk. The platform has suffered dozens of major outages in 2026, and the credibility hit shows up in how developers talk about the service. Mitchell Hashimoto, cofounder of startup HashiCorp, wrote in April that GitHub was “no longer a place for serious work if it just blocks you out for hours per day, every day.” That is the second-order effect boards should care about: reliability issues do not just interrupt work, they can nudge engineers to alternative tools and workflows.

And those alternatives are real within the source’s framing. GitHub had an early lead among engineers, but it has faced more competition lately from AI tools such as Cursor and Anthropic’s Claude Code. The pressure is not only to keep GitHub online. It is also to keep GitHub strategically relevant as AI-assisted coding changes how developers write, test, and deploy code.

The finance side makes the tradeoffs sharper. Microsoft projected that its capital expenditures for the 2026 calendar year will reach $190 billion, largely to expand data center capacity. But the source adds that many data center projects are delayed. Meanwhile, Microsoft has other important AI projects and businesses that must be supported with compute capacity. Add it up and the likely outcome is a multi-pronged plan: buy relief capacity, accelerate the long-term migration to Azure, and in parallel rethink how GitHub competes.

Finally, there is a governance and strategy signal hiding in the weeds. The source says that in an internal meeting late last year, a Microsoft executive spoke about needing to overhaul GitHub to compete with Cursor and Claude Code, according to audio reviewed by Business Insider. That makes today’s AWS capacity move easier to interpret as more than firefighting. It is a way to keep the platform stable while Microsoft executes on a deeper product and infrastructure strategy.

For executives watching from the outside, the takeaway is blunt. When AI demand hits infrastructure like a wave, reliability becomes a competitive weapon, and cloud strategy can’t be purely ideological. Microsoft’s decision to add AWS capacity to stabilize GitHub is a reminder that the fastest way to protect growth is sometimes to accept a messy multi-cloud reality, even when your business model wants the world to be single-cloud.