When I became OSI’s first executive director in 2021, some people thought OSI’s work was done. The Open Source Definition had “won,” embraced even by organizations that once resisted it. Our list of OSI-approved licenses was stable. In their eyes, there was nothing left to do.

I didn’t see it that way.

Technology was (and still is) evolving in ways that test how we apply the Open Source Definition in practice. The launch of the Apple Store for the iPhone for example, made it impossible for users to modify and run an Open Source application on the device (see FSF’s opinion of 2008.) In 2015 smart contracts in the blockchain raised new questions. In October 2021 GitHub announced Copilot, immediately presenting an even bigger challenge: in a world where systems aren’t traditionally “programmed,” what does “source code” mean?

I brought OSI into a leadership role in these conversations, to try to keep them from being pulled apart by competing interests. That’s why we focused on Open Source AI, and why we’ve built OSI into a stronger, more resilient organization ready for the next wave of challenges.

Now it is time to expand OSI’s community and reach out to those unfamiliar with the Open Source principles. We need to apply what we’ve learned about software and technology to the rest of the world where there are many misconceptions about AI and its role in society.  

Defining Open Source AI—Why it matters

One of the most important things we’ve done together is deliver the first Open Source AI Definition. We took this on because we saw a critical gap: without a clear, community-driven definition, multiple versions would inevitably emerge. Policymakers might adopt one—or invent their own—without the input of the Open Source community. That could have chipped away at the Open Source Definition itself, which for decades has been our shared foundation.

From the start, we knew this work had to reflect OSI’s values. We ran the broadest, most inclusive process possible given the urgency of the issue: engaging over 50 global experts, hosting public town halls, and listening to the voices of researchers, developers, corporations, advocacy organizations from across the ecosystem. We did this without the luxury of massive staff, budget, or time, choosing instead to make the best use of what we had to ensure the process stayed open, transparent, and principled.

The result is more than a definition; it’s a testament to the power of community-led collaboration. It proves that the global Open Source community can come together, bridge competing interests, and agree on a shared vision for AI.

For legislators and policymakers, the OSAID provides a clear and trusted framework for navigating complex AI policy debates. It demonstrates that Open Source principles can—and must—guide the development of AI, even in a world of significant technical constraints.

What we built together

Over the past four years, with gratitude to our members, sponsors, volunteers, and partners, we’ve:

Grown OSI’s capacity – Expanded from a staff of one to a professional team with expertise in policy, licensing, communications, and sustainability.

Increased our impact – Engaged policymakers from the U.S. to the European Union to the United Nations, helping shape laws and standards that affect the entire Open Source ecosystem.

Expanded our reach – Grew membership from hundreds to thousands, tripled website visitors, and launched new community resources like the refreshed Opensource.net and member forums.

We also managed to settle the very important debate about the hyphenation in Open Source (it’s never hyphenated and always capitalized, too.)

The work isn’t done

If there’s one message I am taking into my work in the future, it’s this: keep changing.

Open Source is not a finished project. New technologies, regulations, and societal pressures will keep testing our shared principles. New licenses will emerge. New legal and policy challenges will require attention. The OSAID must continue to evolve. But OSI is still a single organization in a relatively small community of dedicated believers. The challenge is moving beyond and outside of that close-knit world to bring an understanding of Open Source for AI to those in leadership positions across industries, governments, and regions. Policies are being written and decisions being made right now that will have widespread impact to those outside of the current community. The Open Source Definition will need ongoing interpretation. The whole concept of “data” needs to be better understood, and why we’re dedicating another Deep Dive to data governance in October (register for the online conference.) 

Everyone from big tech to Melania Trump are trying to make sense of AI and the potential for damage is high. This is where we must now turn our attention.

Thank you, and let’s stay connected

Leading OSI as executive director for a four-year term has centered me in the modern era of Open Source, where regulation has a stronger role for collaboration and impact. I’ll still be part of this movement—Join me in exploring the broader set of challenges in bringing these values to those still unfamiliar with open source and struggling to understand AI. 

Keep up with my new endeavors at maffulli.net.

Let’s keep the conversation going. And above all, let’s keep changing.

Source: opensource.org