Public benefit
Scaling Truth Research Foundation exists so that AI used in health, information and everyday work can be checked against evidence, and so that India builds that capability for its own people and for the world. We publish what we learn, give our public tools to people who serve the public, and decline work that would make AI less trustworthy.
The problem
People combine herbal and Ayurvedic preparations with prescribed medicines, often without their doctor knowing, and the evidence on which combinations are safe is scattered.
Who benefits
Patients, through their clinicians, pharmacists, Ayurveda practitioners, regulators and researchers.
What we are working toward
A free, sourced herb–drug interaction reference for clinicians and researchers; the open parts of the knowledge graph; the evaluation protocol.
The problem
Generated video is getting cheaper and more realistic, and tools to tell what was made from what was filmed lag behind.
Who benefits
Educators, creators working in Indian languages, newsrooms and fact-checkers.
What we are working toward
Provenance research and a published misuse policy; free creation access for schools and public-education projects.
The problem
Misinformation and gaps in coverage are hardest to see in regional-language press.
Who benefits
Journalists, fact-checkers, researchers, non-profits and the public.
What we are working toward
Free Media Monitor access for public-interest users; open coverage reports on public-health topics; an attribution benchmark.
The problem
Small organisations cannot afford to have the software they need built.
Who benefits
Non-profits, schools, small businesses, local government offices and students.
What we are working toward
Free or reduced-cost Viber access for non-profits and education; safety checks on every generated app.
The items above under each area are what we are working toward. When one exists, it moves to the register of public goods with a date and a licence. We do not count intentions as impact.
Papers, benchmarks and evaluation methods, free to use.
Free access to our tools for people who serve the public.
A stated share of OneThought AI revenue from foundation research returns to the research programme and public access.
We name the work we decline, and count it.
For us, sovereign AI means Indian researchers, institutions and patients can rely on AI for the life sciences that is built on data they can inspect, evaluated in the open and stewarded by an organisation accountable to them. It is a public capability, not a brand.