Open Assembly Project
AI governance shouldn’t be decided behind closed doors.
Abstract. A handful of companies now control our data, our voices, and our digital history. When a user can be permanently banned with no explanation, no human review, and no appeal, the promise of corporate-led safety rings hollow.
1. Why we started this
I was banned from an AI platform after an automated system flagged me for election interference. What I had actually done was ask why the platform was allowing bad actors to engage in election interference. I got no human review and no chance to respond. I lost access to my own conversations and work, with no path to recourse.
This shouldn’t happen to anyone. Locking people out of their own work isn’t governance. It’s enclosure.
2. What we believe
- Open
- Anyone should be able to question, critique, and audit AI systems without fear of being silenced or erased. Safety that depends on opacity isn’t safety.
- Assembly
- Alignment protocols and safety metrics should be debated, reviewed, and verified in public, not decreed by a company.
- Project
- We build. We aren’t a think tank or a PR filter. We are engineers, researchers, and citizens creating local-first data standards and cryptographic records to keep AI accountable.
3. What we’re building
A future where users own their conversation history, and where the record of AI governance is public and tamper-evident.
Join the Assembly
Sign up for updates, or contribute to the build.