ROOST is committed to helping accelerate this momentum and building a shared global foundation of open source safeguards that uphold the dignity of every child online.

The movement is already underway. Major platforms have begun releasing their child-safety tooling openly through ROOST, proving it’s both feasible and commercially viable.

This approach empowers families, enables independent auditing, and helps safer tools reach every corner of the ecosystem, not just the largest platforms.

Real accountability and transparency depend on technical openness. That’s why we believe core child-safety protections should be open sourced so anyone can study, test, improve, and reuse them.

One message from Rome stays with us: protecting children online isn’t just about what safeguards exist, but how they’re built.

And save the date! We’re hosting a hackathon on December 8 in San Francisco with OpenAI and Hugging Face. If you’re a developer, Trust & Safety expert, or policy wonk — join us there! events.openai.com/gpt-oss-safe...

If that resonates with you — even if you’re not a developer — there are plenty of ways to join our newly launched ROOST Model Community (RMC):
- Hop into our Discord server (lnkd.in/e4QJNrE4)
- Explore the GitHub repo (lnkd.in/em58E2Ff)

Her words were a reminder that open source AI can only thrive through collaboration — and that building a global community around these principles is how we ensure technology serves the public interest, not the other way around.

Thank you for this amazing write up and the kind words! We will keep doing our best to contribute to and foster a pluralist, resilient open source world!

The full panel is now streaming on YouTube. Make sure to tune in and hear from industry leaders on why this moment marks a milestone for the AI ecosystem — and why ROOST’s work is essential to fostering collaboration and transparency in service of the public good.
www.youtube.com/watch?v=bGKG...

Founding President Camille François shared this sentiment during our panel at the Paris Peace Forum where we announced in partnership with OpenAI the release of gpt-oss-safeguard -- a powerful AI model that can be used to tackle a range of complicated online harms.

@blaine.bsky.social -- we are so honored by your post. 🙏This is just the beginning!

thank you!

And if that doesn’t sound like you, don’t worry – there’s still plenty of ways to join the RMC!

Hop in our Discord server discord.gg/UXmBqy7kFX to learn more.

Or, dive right into the GitHub repo github.com/roostorg/ope....

To kick it off, we’re hosting a hackathon on December 8 in San Francisco with OpenAI and Hugging Face @hf.co . If you’re a developer, Trust & Safety expert, or policy wonk, join us there! events.openai.com/gpt-oss-safe...

Done!