AI Policy
This is a plain-language summary of ImpactMojo's AI policy. The full policy is available at impactmojo.in/ai-policy.html.
Our Position on AI
ImpactMojo uses AI as a tool to support learning — not to replace human teaching, critical thinking, or expert judgement. Every piece of AI-assisted content is reviewed by subject matter experts before it reaches you.
Where We Use AI
AI Study Companions
Each flagship course includes a study companion powered by Google's NotebookLM
These help you explore course content through conversation. They're trained on course materials, not the open internet.
Content Development
AI assists with research synthesis and initial content drafting
All AI-drafted content is verified by Dr. Varna Sri Raman and the team against peer-reviewed sources before publication.
VaniScribe
AI-powered transcription of field interviews in South Asian languages
Uses Sarvam AI for language processing. Your audio is processed and not stored permanently.
Qual Insights
AI-suggested codes for qualitative analysis
Suggestions are starting points — you review and decide what to accept.
Translation
AI assists with translating content into the 6 supported languages
All translations are reviewed for accuracy and cultural appropriateness.
Where We Do NOT Use AI
These are things ImpactMojo explicitly prohibits:
Publishing unverified statistics. AI can generate plausible-sounding numbers that are completely wrong. We never publish AI-generated data without verification against primary sources.
Generating fake citations. Every paper, study, and source we reference is real and has been checked.
Making decisions about learners automatically. Your progress, certificates, and access are never determined by AI systems.
Training AI models on your data. Your learning activity, notes, bookmarks, and any content you create on the platform is never used to train AI models.
Deepfakes or impersonation. We never use AI to generate synthetic voices, images, or personas that pretend to be real people.
Replacing instructor judgement. AI doesn't grade your work, assess your competence, or make recommendations about your career.
Our Quality Control Process
When AI is involved in content creation, it goes through a multi-stage review:
Subject matter expert review — Does the content accurately reflect the evidence?
Fact-checking — Are statistics, citations, and claims verified against primary sources?
Context review — Is the content appropriate and sensitive to South Asian development contexts?
Founder approval — Final sign-off by Dr. Varna Sri Raman
What About Mojini?
Mojini, ImpactMojo's platform assistant chatbot, is not AI-powered. It's a structured FAQ system that matches your question to pre-written answers. It doesn't use a large language model, doesn't generate text, and doesn't learn from your conversations. See the Mojini Guide for details.
How Often We Review This Policy
The AI policy is reviewed quarterly (every 3 months). When changes are made, they're communicated via the ImpactMojo newsletter and posted on the website.
Questions?
If you have questions about how ImpactMojo uses AI, email hello@impactmojo.in. We'll give you a straight answer.
Last updated
Was this helpful?