Ubuntu — “I am because we are” — is not a platitude. It is a philosophical framework that has governed human relationships across southern and eastern Africa for centuries. It describes a fundamentally different model of personhood, one in which identity is constituted through community rather than in isolation from it.
I have been thinking about what Ubuntu means for AI governance since building LadenX, an autonomous AI agent that manages production servers. The system's approval gates — where the AI proposes an action and a human must approve it before execution — were a practical engineering decision. The AI shouldn't act alone because the consequences of autonomous action on production infrastructure are too significant for any single entity, human or artificial, to bear alone.
But when I examined what I'd built through the lens of Ubuntu, I realized the approval gates embody something deeper than risk management. They encode a communal model of accountability.
Individual vs. Communal Accountability
Western AI governance frameworks tend to focus on individual responsibility. Who trained the model? Who deployed it? Who pressed the button? The EU AI Act, for example, assigns risk categories and compliance requirements to specific entities — the provider, the deployer, the user. Accountability is traced along a chain to an individual decision-maker.
This model has value, but it has a blind spot. AI systems are not the product of individual decisions. They emerge from collective processes — teams that build them, organizations that deploy them, societies that adopt them, and communities that are affected by them. Assigning accountability to a single entity in that chain obscures the collective nature of the technology.
Ubuntu offers an alternative. In Ubuntu philosophy, accountability is distributed. The community is responsible for its members' actions, and members are responsible to the community. No one acts in isolation, and no one bears consequences alone.
Ubuntu in Code
LadenX's governance model works this way. The AI doesn't act alone — it operates within a community of human overseers. When it proposes a dangerous operation, the approval isn't a single person's individual decision. It's a checkpoint within a system where responsibility is shared across the team that built, deployed, and operates the agent. The audit logs don't just track what happened — they create transparency that makes communal oversight possible.
This isn't theoretical. At Command Space, where I serve as Technical Director with a team of seven, we operate on principles that are fundamentally communal. Decisions about how our AI tools operate aren't made unilaterally. At Girl Code Africa, where I serve as CTO and mentor over 20 developers, the approach is even more explicitly community-centered. We build capacity together. We learn together. We are accountable to one another.
Beyond the Global North
The current AI governance conversation is dominated by perspectives from the Global North. The major frameworks — the EU AI Act, the NIST AI Risk Management Framework, China's regulatory approach — reflect the philosophical traditions of the societies that produced them. These are valuable contributions, but they are not the only contributions possible.
Africa's philosophical traditions offer frameworks for thinking about collective responsibility that the current governance conversation needs. Ubuntu is not the only one — concepts of communal decision-making, elder oversight, and distributed accountability exist across the continent's diverse philosophical traditions.
Ubuntu Questions
When I facilitated AI workshops for the Stanford SEED Transformation Network in Accra in August 2025, with roughly 150 entrepreneurs and executives in the room, the conversations about AI ethics naturally gravitated toward communal concerns. The questions weren't “How do I protect my individual data?” They were “How does this technology affect my community? Who in my community should be involved in deciding how we use it? What responsibility do I have to the people around me if I deploy this?”
These are Ubuntu questions. They deserve Ubuntu-informed answers.
I am not arguing that Ubuntu should replace existing governance frameworks. I am arguing that it should inform them. The current frameworks are incomplete without perspectives from communities that understand accountability as a shared rather than individual construct.
Africa is not just a market for AI products developed elsewhere. It is not just a population to be governed by frameworks written elsewhere. Africa has philosophical traditions that can make AI governance more robust, more equitable, and more human.
The approval gate in LadenX is a small piece of code. But the principle it embodies — that no intelligent system should act without the consent of the community it serves — is as old as the continent I build from.