A public that can't see how a system works can't hold it accountable.
The most advanced technology in public life is often the least understood by the people it affects. We change that — translating systems into plain language, building transparency into the open, and advocating for technology that serves people first.
A discipline, not a press release.
Accountability isn't possible without understanding. When a benefits algorithm or a data system is a black box, the public can't consent to it, question it, or push back — and harm hides in the gap. Informing the public isn't marketing; it's the safeguard that makes every other safeguard real. So we treat it like engineering: a deliverable, with a standard, that ships alongside the build.
Transparency is about understanding — not exposure.
We make how a system works legible to the people it affects — what it does, what data it touches, and where it can go wrong. Many systems still need logins, role-based access, and strict limits to protect people; transparency and access control aren't opposites. And where a system reports results to the public, we hold a higher bar.
Example · where a system reports to the public
Plain language · public results
Same published results · no privileged version
Our transparency commitments
One truth for everyone.
Where a system reports to the public, residents see what officials see — no privileged dashboard, no watered-down version.
Show the blind spots on purpose.
Every result ships with its limits — who was heard, who wasn't, and how confident we are.
Plain language first.
If a community can't understand it, we haven't finished explaining it.
Prove it, don't promise it.
Protections you can verify beat assurances you take on faith.
The public record.
Plain-language explainers and open research — part of how we keep the conversation honest.
One truth for everyone
Why the public should see exactly what officials see.
February 2026 · 4 min ExplainerHow to read an algorithm deciding your benefits
A plain-language guide — five questions to start with.
March 2026 · 6 min ApproachBeyond AI
The four things that actually decide whether a system serves people.
April 2026 · 6 min"Sunlight isn't a feature we add at the end. It's how the public keeps the rest of us honest."