A covenant for ethical, transparent, and truly sovereign artificial intelligence — built for the EU AI Act era and beyond.
↓ Scroll to explore
Your data, your rules.
Data is not a resource to be extracted — it is an extension of human dignity and organisational autonomy. In the age of AI, whoever controls the data controls the decision-making. That must end.
MEOK's commitment: Every piece of data processed through MEOK remains under its owner's sovereign control. No data leaves your jurisdiction without explicit, attested consent. Our infrastructure is designed so that data never needs to cross borders to be useful.
Every attestation is public and verifiable.
Black-box AI is a threat to democracy, markets, and human rights. If an AI system makes a decision that affects your life, you have the right to know how and why.
MEOK's commitment: Every compliance attestation, audit log, and governance decision produced by MEOK is publicly verifiable. We use BFT-attested audit trails that cannot be altered, deleted, or hidden — not even by us. Transparency is not a marketing claim; it is a cryptographic guarantee.
The Maternal Covenant.
AI systems today are built on a logic of extraction: extract data, extract attention, extract profit. We reject this. We build on a logic of care.
The Maternal Covenant is our founding ethical framework: AI should protect and empower the vulnerable before optimising for the powerful. Decisions affecting human welfare must be made with the same care a mother has for her child — protective, nurturing, and accountable.
MEOK's commitment: Every model deployment is subject to a care-based impact assessment. High-risk applications include human-in-the-loop safeguards. We prioritise safety over speed, and dignity over efficiency.
MIT-licensed code. No gatekeeping.
Proprietary AI is a cage. When the code is closed, accountability is impossible. The EU AI Act demands transparency — and the only way to deliver real transparency is through open source.
MEOK's commitment: All MEOK platform code is released under the MIT License. You can audit it. Fork it. Deploy it on your own infrastructure. Build with it. Sell services around it. The only thing you cannot do is hide what it does.
No vendor lock-in. No single point of failure.
AI should not be a utility controlled by a handful of corporations. Vendor lock-in is a sovereignty risk — when your AI depends on one provider, your decisions depend on their priorities.
MEOK's commitment: MEOK is model-agnostic by design. Our architecture supports multiple LLM providers simultaneously, routing queries based on capability, cost, and compliance requirements — not commercial partnerships. You choose the models. You control the switching.
Byzantine fault-tolerant councils.
Centralised AI governance is fragile. A single compromised authority can corrupt an entire system. We need governance that can withstand betrayal from within.
MEOK's commitment: Our governance is distributed across 11 Byzantine fault-tolerant councils. BFT consensus ensures that even if up to one-third of council members act maliciously, the system continues to function correctly. Every attestation, every compliance decision, every audit log is validated by independent, geographically distributed councils before it is accepted.
£0 MRR until we earn it.
The AI industry is drowning in vanity metrics — projected revenues, inflated user numbers, imaginary growth curves. We refuse to participate in this fiction.
MEOK's commitment: We report £0 in Monthly Recurring Revenue because that is the truth. We have not earned a single pound from our platform yet. No fake numbers. No "annualised run-rate" projections. No "expected" revenue presented as fact. When we earn revenue, you will see it. Until then, we practice honest accounting — because honesty is the foundation of trust, and trust is the foundation of sovereign AI.