Audiopedia Foundation

SILAC

Community-controlled AI for Indigenous communicators

SILAC - Sistema Local para la Adaptación de Contenidos - is an approach to using AI and external knowledge on community terms.

The idea

AI should help Indigenous communicators access and adapt useful knowledge without requiring communities to surrender control over their own languages, knowledge or context.

How it works

SILAC combines selected trusted sources, locally governed knowledge and AI-assisted adaptation in a workflow controlled by local communicators.

AI can help search, compare, summarize, translate and prepare draft content.

People decide what sources are used, what knowledge enters the system, what is appropriate to combine, and what is published.

PriNciples

Community control

Communities define what the system may use and how.

Human judgment

AI produces drafts, not authority.

Provenance

Sources should remain visible and traceable.

Local knowledge governance

Not all knowledge should enter an AI system.

Local-first where possible

Systems should work with limited connectivity and minimize unnecessary external data flows.

Open and replaceable technology

SILAC should not depend on a single model, vendor or platform.

Why

The goal is not to bring more Indigenous knowledge into AI.

The goal is to help Indigenous communicators use AI and external knowledge on their own terms.

Current work

SILAC is being developed as part of Indigenous Audio Action.

Initial co-development is being explored with Indigenous communicators and partners in Peru, with a focus on practical workflows, governance and locally useful communication.

Status
In development.

The approach will evolve through real community use rather than being designed as a finished technology in advance.

GET INVOLVED

SILAC is still taking shape.

We welcome conversations with Indigenous communicators, researchers, funders and technologists exploring community-controlled AI, knowledge governance and local-first systems.

GET IN TOUCH
SILAC is developed as part of Indigenous Audio Action, an initiative of Audiopedia Foundation.

© 2026 Audiopedia Foundation