The License Change That Started Everything
In August of 2023, HashiCorp made a decision that reshaped the infrastructure-as-code landscape. They moved Terraform from the Mozilla Public License to the Business Source License, a move that sent ripples through every organization running Terraform at scale. I remember reading the announcement and immediately understanding what was coming. The BSL model means the code is open and readable, but commercial use requires a license after a certain threshold. For companies, this created genuine uncertainty about their existing deployments and future commitments.
The Linux Foundation didn’t wait long to act. OpenTofu emerged as a fork, maintained under governance that promised to keep the project genuinely open. It was a fork born not from technical disagreement but from philosophical and business model concerns. At the time, many dismissed it as a reaction that wouldn’t survive long term. Forks often don’t. This one did something different.
Two Years of Divergence: Feature Parity Shifts
What matters most when evaluating these tools in 2026 isn’t the licensing debate anymore. The technical landscape has shifted beneath that argument. OpenTofu released version 1.9 in late 2025, and this version introduced capabilities that remain absent from Terraform’s corresponding releases. Provider-defined functions arrived first in OpenTofu, giving providers the ability to expose custom logic directly to your configuration. State encryption, a feature that should have been table stakes years ago, also shipped in OpenTofu ahead of its Terraform equivalent.
This matters more than people think. When a fork starts shipping features faster than the original, it signals something real about momentum and prioritization. I’ve watched enough projects to know the difference between a temporary lead and a structural advantage. OpenTofu’s official documentation and changelog shows commit frequency that exceeded HashiCorp’s Terraform in several consecutive months through 2025. That’s not noise. That’s an active community making deliberate choices about where to invest engineering effort.
The IBM Acquisition and Enterprise Uncertainty
Then came June 2024. IBM acquired HashiCorp for 6.4 billion dollars. I need to be careful here because I’m not interested in speculation, but facts are facts. Enterprise customers began asking pointed questions about Terraform’s trajectory under a company whose core business is very different from HashiCorp’s infrastructure automation focus. IBM’s subsequent communications about Terraform’s roadmap introduced exactly the kind of uncertainty that the Pulumi 2025 survey later quantified: 41% of platform engineers cited IaC tool licensing and governance uncertainty as a top-three risk factor in their 2025 and 2026 planning decisions.
This is the part that keeps infrastructure teams awake at night. It’s not theoretical. When your infrastructure-as-code tool lives inside a 6.4 billion dollar acquisition, the calculus changes. What happens in three years when the quarterly targets shift? I’ve seen enough acquisitions to know this is a legitimate concern, not paranoia.
The Migration Numbers Tell Their Own Story
By early 2026, OpenTofu’s provider registry had mirrored over 2,000 providers. The GitHub repository crossed 23,000 stars. These are not vanity metrics. The provider mirror milestone means you can actually use OpenTofu for most real-world infrastructure problems. Terraform has an ecosystem advantage, but OpenTofu has closed that gap enough to matter for the majority of deployments.
A Spacelift survey conducted in 2025 found that 34% of infrastructure teams actively evaluating OpenTofu had completed a full migration from Terraform. Another 28% reported partial migrations in progress. That’s 62% of evaluating teams who have moved at least some workloads. These aren’t small numbers, and they’re not coming from bleeding-edge early adopters anymore. These are teams making deliberate choices based on concrete factors.
The Pulumi State of Cloud Infrastructure 2025 report adds useful context. Teams are diversifying their IaC investments specifically because of licensing uncertainty. When 41% of platform engineers cite governance and licensing as top risks, it’s not a marginal concern. It’s a central planning factor.
Making the Call in 2026
So where does this leave you if you’re trying to make a decision today? Honestly, it depends on your specific constraints in ways that would have been unthinkable in 2023. If you have deep investment in Terraform, vendor lock-in considerations are real, but so is stability. The tool still works. It will continue to work. HashiCorp, now under IBM, isn’t going to let Terraform collapse. But you should go into that choice with eyes open about the licensing model and governance questions.
If you’re starting something new or evaluating migrations, OpenTofu is a genuine alternative that has matured past the fork stage. The feature parity question is mostly settled now. The ecosystem is solid. The governance is transparent. The community velocity is real. The decision is no longer obvious, which is exactly what you should expect from a fork that actually mattered.
The irony is that HashiCorp’s license change inadvertently did something useful for the infrastructure community. It forced a conversation about tool governance that should have happened earlier, and created real incentives for an alternative that has turned into something legitimate. If you haven’t looked seriously at OpenTofu since 2023 or 2024, the project that exists in early 2026 is worth your time.