ROOST
@roost
We’re ROOST (Robust Open Online Safety Tools), a non-profit building and maintaining open source, accessible, high-quality, transparent safety tools for digital organizations of all kinds. roost.tools
ROOST OSS Community Manager @cassidyjames.com is at #FOSSY in Vancouver through Sunday! FOSSY runs Aug 6–9 at UBC, hosted by the Software Freedom Conservancy. Our tools are free, open source, and self-hostable. The latest Osprey ships ATProto plugins. Come say hi! roost.tools/blog/osprey-...
ROOSTERs @julietshen.bsky.social & @kristinalvandi.bsky.social are at the Crimes Against Children Conference #CACC2026. We work w/everyone (platforms, nonprofits, vendors) to make detection, review, & reporting tools widely available so all organizations can study, modify, & use them. Come see us!
That's a wrap on TrustCon 🐓 3 days, 6 sessions, 1 launch (Osprey 1.1.0!), many chicken hats. Favorite moment: non-engineers building working moderation rules in our workshop — safety tooling you can study, modify, and use. One year of Coop & Osprey. Just getting started.
TrustCon day 2 💚 Thank you to everyone who packed yesterday's panel on what open source means for your T&S stack with @julietshen.bsky.social. Fitting week: 1 yr ago today we introduced Coop & Osprey. Today they help protect 360M+ users, including 45M events/day right here on Bluesky. roost.tools
We celebrated last night by convening fellow allies in Geneva for a conversation on building AI safety in the open and heard from builders wanting better safeguards for kids-focused products, from funders architecting public interest infrastructure, and from researchers working on AI accountability.
On the integration side, Coop now supports OpenAI's omni-moderation-latest model as an image moderation signal source. HMA (hasher-matcher-actioner) exchanges can now be configured directly from the Coop UI, removing the need to manage that setup separately. You still need your own NCMEC creds!
We had multiple platforms adopt Coop during the 1.0 development cycle, which gave us critical real-world feedback. Several significant capabilities shipped as a result: User Strikes are now enabled, giving platforms a native way to configure, track, and act on repeat violations directly within Coop
Coop 1.0 comes with a simplified infrastructure and ways to deploy, new features, and (much) improved documentation! Check out what's possible for your team and explore the features here: www.loom.com/share/3043f9...
En matière d'IA et de sécurité en ligne, la protection des enfants est une priorité absolue. Aujourd'hui, ROOST annonce, à l'occasion du G7 à Paris, un partenariat avec @mila-quebec.bsky.social afin d'intégrer des recherches de pointe à nos efforts pour développer des outils en source ouverte.
If you see this, I implore you to post a picture from whatever device you're using with no explanation.
say hello to the ROOST team if you're at #atmosphereconf @julietshen.bsky.social @cassidyjames.com
join @julietshen.bsky.social in room 2311 for an Unconference session on trust and safety at 3:30! #ATmosphereconf
We're on the ground at GDC Festival of Gaming this year! Let us know if you're there, too, and come say hi to our COO @annebdh.bsky.social. And if you get a hold of one of our special, limited postcards, take a photo and share it with us!
On Monday, @hf.co , OpenAI, and ROOST hosted the first Open Safeguard Hackathon. The room was filled with people who care deeply about safety, working together to explore and test gpt-oss-safeguard. Check out the ROOST Model Community for some of the projects github.com/roostorg/mod...
This week, our President, @camillefrancois.bsky.social, visited the Vatican for a discussion on protecting children in the digital age. We were very humbled to have been invited to participate in such an important meeting and at such a pivotal time.
During our @parispeaceforum.bsky.social panel last week, Dr. Giada Pistilli eloquently explained that transparency in AI isn’t just a technical challenge — it’s a societal one. The more people who can see, shape, and question these systems, the more accountable and representative they become.
Thank you to Martin Tisné, Chair of the Board at Current AI and CEO of the AI Collaborative, for speaking on our panel this week at the Paris Peace Forum and for your contributions towards this work in open source and AI.
Thank you to Fidji Simo for addressing the audience via video yesterday ahead of our panel at the Paris Peace Forum. The conversation on how AI systems can help create a safer internet for the public good through open source tooling is only just beginning. lnkd.in/e_jkiemB
Excited to announce here at the @parispeaceforum.bsky.social that, after months of joint technical work together, OpenAI will open source through ROOST one of its main internal models that handles detection and moderation. This will allow anyone or any company to use, test, and study the model.
Looking forward to a great panel tomorrow at the @parispeaceforum.bsky.social Forum, hosted by ROOST President, @camillefrancois.bsky.social. Stay tuned for more! #onlinesafety #AI #AItransparency #opensource #trustandsafety #techforgood
Our Head of Product, @julietshen.bsky.social, did a live demo of our Osprey tool, and led a feedback session with fellow ROOSTer Felix Wang at this official side event alongside the UN General Assembly. We will be digitizing and publishing this feedback under the Chatham House Rules soon!
If you're at the Trust & Safety Research Conference, check out the @roost.tools and @discord.com workshop on open source happening at 11:30am PT TODAY! Fisher Conference Center, Barnes/McDowell/Cranston #TSRConf #TrustAndSafety
We’re proud to carry forward the legacy of Cove. Their work helped report 120K+ cases of child exploitation — and now their IP will live on as open source, stewarded by ROOST. Learn more about the announcement of our new tool “Coop” below: roost.tools/blog/roost-a... #TrustAndSafety #OpenSource
#ICYMI: ROOST is at TrustCon 2025! We kicked off the conference with a keynote address from our CTO, Vinay Rao, who shared that a safer digital landscape can be achieved through increased awareness, community engagement and technological advancements. #TrustandSafety #TrustCon