Happy Saturday,
Every building has plumbing. Yours too. You don’t think about it until something breaks. Then nothing works.
The BBNJ Agreement has plumbing too. It’s called the Clearing-House Mechanism. The CHM. And during PrepCom 2, we’ve been going back and forth on how to build it.
Here’s what the CHM actually does: It controls how data moves through this treaty. Who submits what. Who gets to see what. And what happens when a researcher’s work touches your Traditional Knowledge.
Get the plumbing wrong?
Every protection we fought for leaks out.
Last year at PC2 nobody could agree. The co-chairs brought in consultants.
They prepared a study and came back with three options.
Let me make these dead simple.
Option A. One big filing cabinet. Secretariat runs everything. One database. One system. Fastest to build. 9 to 12 months. 2 to 4 million bucks. But if you’re sitting on a Pacific island with spotty wifi, you’re sending your data across the world and crossing your fingers.
Option B. Same filing cabinet, but now you’ve got local offices too. Regional hubs that help communities prep submissions, work offline when the internet drops, sync when they reconnect. More moving parts. But the Traditional Knowledge stays near the people it belongs to. That matters.
Option C. Grab something old and bolt BBNJ onto it. Take the Nagoya Protocol registries or something similar and retrofit them. Sounds smart. Right?
I’ve seen this movie before.
Under the UNFCCC Article 6.4 carbon market negotiations, the exact same thing happened. Old credits from the Clean Development Mechanism, pre-Paris Agreement, no safeguards, and people wanted to carry them straight into the new system. Laundering dirty credits into a clean market.
Option C is the data version of that. Old forms from a system that never had to meet BBNJ standards poured into the new treaty’s pipes. Different treaty. Same trap.
So we said: combine A and B. Skip C.
Build the central safety net. Pair it with regional hubs. And when it comes to Traditional Knowledge, the system signals that the knowledge exists. It never forces you to upload the knowledge itself.
Think library catalog. The catalog tells you a book exists and where to find it. It doesn’t take your book off your shelf and put it on someone else’s.
Now here’s the part that decides whether this system protects you or screws you.
The consultants list nine design choices the Conference of the Parties has to make. Here’s what we put on the floor.
The second a researcher flags Traditional Knowledge involvement, the identifier flips to “Pending Consent.” Everything freezes. Nothing moves until that researcher gets real consent from the actual Indigenous Peoples involved and uploads a digital receipt proving it happened.
Not a checkbox.
A real document. Verified. Off-platform.
We proposed a “red light default.” The identifier starts frozen and stays frozen until off-platform consent is verified. Red light until proven green.
And the system needs offline submission with queued sync. Build a system that only works on fast internet and you just locked out the regions and communities it was built to protect.
Then there’s confidentiality.
Most people get this one wrong. Labeling something “restricted” on a central server does not protect it. You moved someone’s knowledge to a third-party server and stuck a note on it that says “please don’t look.”
That’s not protection. That’s a polite suggestion.
The baseline has to be signaling, not transferring. The knowledge never leaves the jurisdiction of the rights-holders. The system says “this exists.” The actual knowledge stays where it belongs.
User roles sound technical. They’re not.
The system needs a category for Indigenous Peoples that actually recognizes who you are. Legacy databases either leave us out, lump us in with “civil society,” or use State governments as our consent proxies.
Every single one of those breaks FPIC before the system even goes live. If the user taxonomy doesn’t treat Indigenous Peoples as a distinct category, the consent architecture is dead on arrival.
And look, the system has to reject all uploads of substantive Traditional Knowledge.
No narratives. No files. It’s a metadata routing directory. Access pathways go directly to designated Indigenous Peoples, their reps, their institutions. Not through State National Focal Points. Direct.
Interoperability loops right back to Option C.
Moving Traditional Knowledge from old databases into the new system without a consent audit? That’s data laundering. No automatic grandfathering.
Any legacy Traditional Knowledge data that wants a BBNJ identifier goes through a methodology designed by the proposed Indigenous Advisory Mechanism first. Meet the current FPIC baseline or no identifier. No exceptions.
On cybersecurity, we pushed hard-coded boundaries against AI. Automated scraping, large language model ingestion, algorithmic training on Indigenous metadata. Prohibited. Audited.
Let AI tools ingest Indigenous metadata without consent and you just built a pipeline that converts Traditional Knowledge into training data. That’s digital biopiracy.
Here’s what all of this adds up to.
The Indigenous Advisory Mechanism tells the CHM what counts as valid consent. Not the administrators. Not the Secretariat. The Indigenous body writes the rules. The pipeline freezes until consent protocols are met. Only the Indigenous body decides what “met” means.
The CHM cannot function without the Indigenous Advisory Mechanism. Pipeline stops. Identifier stays grey. Nothing moves until the Indigenous body says it moves.
That’s what one of the things I conveyed in under 5 minutes this week.
This is a bit of divergence from your regular newsletter.
Back to our regular programming next week!
P.S.: Please share this with your friends.
P.P.S: This week I uploaded 4,5 hours of Indigenous diplomacy briefs, a snapshot of what the first week of BBNJ PrepCom 3 covered:
I showed how I covered two parallel negotiations using budget diplomacy.
A co-chair opened the floor on a topic I had no prepared intervention for. I showed what I did in real time and why being deep enough in the substance lets you work from headlines.
I broke down how I answer member state questions inside my own intervention. I gave a masterclass in 2,5 minutes. No history lesson. No Declaration lecture.
I recognized a pattern from Article 6.2 carbon market negotiations inside the clearinghouse mechanism discussion. Same architecture question. Different treaty. That pattern unlocked a proposal nobody else was making.
I started designing a structural fix for conflation that does not depend on language.
I taught the GPS framework to a cohort of Indigenous women. No slides. One hour. I showed where Indigenous women can have direct impact on processes most people do not even know exist.
The UN80 resolution was adopted by the General Assembly after Russia tried to stop it. I broke down what phase two means and what will happen when Indigenous Peoples will walk into the UNPFII next month.
I mapped three layers that need protecting in UN80. Most people are focused on the bottom layer. I explained why that is the wrong order.
