Ethical framework.
Seven principles govern every piece of research, policy, and technology work Neuravox Foundation undertakes. They bind our staff and partners alike.
Last updated: July 2026
This framework applies to all work carried out by or in the name of Neuravox Foundation, including commissioned research, collaborative projects, and community programmes. Where local law, funder requirements, or partner policies conflict with these principles, the stricter standard applies. Concerns about our adherence to this framework may be raised, including anonymously, at contact@neuravox.org.
1. Human agency and oversight
Technology should widen human choice. Every system we study, build, or recommend must remain subject to meaningful human control, and people affected by automated systems, AI among them, must retain the ability to question, contest, and override those systems.
- We assess every project for its effect on the autonomy of the people it touches.
- We recommend human review for any consequential decision informed by an automated system.
- We reject work that treats people as passive subjects of technology.
2. Community consent and data sovereignty
Communities own their stories, their knowledge, and their data. Research conducted with communities in Africa and the Global South must begin with their informed consent and remain accountable to them throughout.
- We seek free, prior, and informed consent before collecting data from or about communities.
- We negotiate how data will be stored, used, shared, and retired before work begins.
- We support community control over data about communities, including the right to withdraw it.
3. Transparency and explainability
People are entitled to understand the systems that affect them and the research that claims to speak about them. We work in the open wherever safety and consent allow.
- We publish our methods, funding sources, and material assumptions alongside our findings.
- We explain technical work in language accessible to the communities it concerns.
- We disclose the limitations and uncertainty of our conclusions.
4. Do no harm
The burden is on us to anticipate harm before it happens. We weigh the risks of our research and recommendations against their benefits, with particular attention to people with the least power to absorb harm.
- We conduct harm and risk assessments before beginning fieldwork or publishing sensitive findings.
- We decline work whose likely harms outweigh its public benefit, regardless of funding.
- We monitor the downstream use of our work and correct or retract it when it is misused or found wrong.
5. Equity and inclusion
The governance of technology has too often been written without the regions most affected by it. We centre voices from Africa and the Global South, and within them the people most often excluded: women, young people, rural communities, and people with disabilities.
- We design research agendas together with underrepresented communities.
- We compensate community contributors fairly for their time and expertise.
- We publish and convene in ways that lower barriers of language, cost, and connectivity.
6. Accountability
Principles mean little without mechanisms. We hold ourselves answerable to the communities we serve, to our partners, and to the public, and we expect the same of the institutions we study.
- We maintain clear channels for complaints and concerns, reachable at contact@neuravox.org.
- We review our own compliance with this framework annually.
- We name a responsible owner for every project, so accountability rests with a person.
7. Open knowledge
Knowledge produced in the public interest belongs to the public. We treat openness as the default and closure as the exception that must be justified.
- We publish research openly and free of charge wherever consent and safety permit.
- We license published work for reuse with attribution, so others can build on it.
- We share tools, datasets, and methods that help others do public interest work of their own.