Why We Open-Sourced Our AI Widget Under MIT
Why we open-sourced the AIML.chat widget under MIT: trust through transparency, no lock-in for customers, and distribution through a clean embed.
We released the AIML.chat widget as open source under the MIT license because the code that runs on our customers' sites should be inspectable, because customers should never be locked into a black box, and because a permissive, trustworthy embed is itself a distribution advantage. Open-sourcing the widget was a strategic choice, not a giveaway.
The widget runs on someone else's site
The widget is the one piece of our product that executes on a customer's web pages, in front of their visitors. That makes it different from our backend: it is code we are asking site owners to trust with their users and their brand. The honest way to earn that trust is to make the code readable — anyone can see exactly what the widget does, what it loads, and what it sends, with no hidden behavior. For something that runs in your visitors' browsers, "trust us" is a weaker promise than "read it yourself," and MIT lets us make the stronger one.
No lock-in, by design
info
MIT is maximally permissive: anyone can use, modify, and self-host the widget. A customer is never trapped — if they want to host the embed themselves or fork it, they can, which means they stay because the product is good, not because leaving is hard.
We chose the most permissive common license deliberately. A restrictive or copyleft license would add friction for exactly the developers and businesses we want to adopt the widget, and it would undercut the no-lock-in promise. MIT removes the legal friction entirely: companies with strict policies can use it, developers can fork it, and no one has to worry about license terms biting them later. The result is a customer relationship built on the product continuing to earn its place, which is the only kind worth having.
Open source as distribution
Beyond trust, open-sourcing the widget is a growth channel. A public, MIT-licensed repository is discoverable, linkable, and forkable; developers find it, try it, and share it in ways a closed product never travels. It signals that we build in the open and play well with the open-source ecosystem our buyers already live in, and it lets integrations, examples, and community contributions accumulate around the widget. For a young product, that kind of organic, credibility-building distribution is worth more than the marginal protection a closed license would offer.
Where we draw the line
Open-sourcing the widget does not mean open-sourcing everything. The widget is the client that runs on customer sites; the backend — ingestion, retrieval, the agent orchestration, billing — remains our hosted service, which is where the operational value and the business live. That split is intentional: make the part that needs trust transparent and the part that needs to be a sustainable service hosted. It gives customers the inspectability and freedom that matter most to them, while keeping the company viable — and we think that balance is exactly right for an infrastructure product.
Getting started
Use the open-source embed on the AIML.chat free plan: index your content and drop the MIT-licensed widget on your site. Read how to add AI to your website in 2026, building an AI agent platform on .NET 10, and compare tiers on the pricing page.
Why open-source the widget under MIT specifically?+
MIT is maximally permissive, so any developer or business can use, modify, or self-host it without legal friction — which delivers the no-lock-in promise the widget is meant to make.
Does open-sourcing the widget mean the whole product is open?+
No — the widget runs on customer sites and needs to be trustworthy and inspectable. The backend remains a hosted service, which is where the operational value lives.
How is open source a distribution advantage?+
A public, forkable repository is discoverable and shareable in ways a closed product is not, building credibility and organic reach within the ecosystem our buyers already use.
Related reading
Get the AIML.chat newsletter
Practical tips on AI docs assistants, RAG, and growing with content — a couple of emails a month. No spam, unsubscribe anytime.
Double opt-in — we'll email you to confirm.