A few weeks ago, Tagoross published the first open content package: AI Openness: Terminology and Definitions guide. It was an attempt to bring some order to a confusing vocabulary. But there was another question behind it.
How did we get here?
To answer it, we at Tagoross put the terms on a timeline. The result is the second Tagoross open content (CC license) package: AI Openness: Terminology Timeline.
The timeline starts in 1986, with the Free Software Definition. It continues through the Open Source Definition in 1998 and the Open (data) Definition in 2006. Then the pace changes. Foundation models appear in 2021. Frontier models follow in 2023. Llama 2 is released in July 2023 and described as open source, while the meaning of open source AI is still being debated.
In 2024, things accelerate again. New ways to define and measure AI openness appear. The EU AI Act also introduces new legal terminology. By 2025 and 2026, the vocabulary has grown further. Open weights, AI openness, digital sovereignty, open washing and other terms are now part of the conversation.
Looking at them together helped us at Tagoross to see something important.
There is not one conversation about AI openness.
There are different conversations about availability, access, transparency, rights, capabilities, public value, sovereignty and business. And many of them use the same word: open. This is why history matters.
The timeline does not try to decide which definition is right. It shows where the terms came from and how the language evolved. It is a snapshot, complementing the guide. The timeline ends in September 2026. The terminology, licenses and regulations will evolve. So the AI Openness resources Tagoross has created.
The question to ask is simple: When someone says that an AI model is open, what exactly do they mean? What implications does it have in relation with vendor lock-in risks and digital sovereignty?
You can explore the timeline and download the second Tagoross open content package here:
AI Openness: Terminology Timeline — Tagoross
Read about the AI Openness: Terminology and Definitions guide in a previous blog post.
