Here's WALDO: AI's first open source project and community

Most sovereignty conversations stop at infrastructure: where a model runs, who hosts it, who can see your prompts. That covers three of the four things an organization needs to control. It skips the first and most basic one: the model itself, and whether anyone can verify what actually trained it.

Join Gregory Kurtzer, founder and CEO of CIQ and founder of OpenWALDO and Rocky Linux, and Addison Snell, CEO of Intersect360 Research, for a conversation grounded in two new pieces of research and reporting: Intersect360's enterprise AI sovereignty whitepaper, and the launch of OpenWALDO, an open source project building a shared, community governed corpus of AI training data.

Open weight models solved part of the sovereignty problem: you can download a model and run it on your own terms. But open weight is not open source, and enterprises are increasingly asking the harder question underneath: what data trained this model, under what license, and can I actually check? Kurtzer and Snell will walk through the sovereignty framework enterprises need to apply to AI, where open weights fit and where they fall short, and what a truly open, verifiable foundation for AI looks like in practice.

• Why sovereignty has to extend past infrastructure and into the model itself

• The difference between open weight and open source, and why enterprises should care

• What an AI Bill of Materials is, and how it lets you verify what trained a model

• How OpenWALDO's community governed corpus gives labs, enterprises, and researchers a verifiable baseline to build on

• What the Intersect360 survey data shows about how enterprises are actually thinking about AI control today