On the eve of the AI Impact summit in India in February, it is clear that most countries still do not have a workable model for managing this technology. The US has pretty much left the issue to market forces, the EU relies on regulatory oversight, and China relies on the concentration of power with the state. But none of these options are realistic if your country, like many others, has to manage AI without large regulatory structures or massive computer power. No, we need a different system that embeds the principles of transparency, consent and accountability directly into the digital infrastructure.