Research sprint · Saturday, 5 December
Building the AI safety ecosystem in India.
We train researchers, run frontier AI safety courses, and build direct talent pipelines from India to global alignment labs.
Up next · in 19 days
Threat Intelligence & Threat Modelling Hackathon
Saturday, 5 December, Saturday · 10:00 am · Bengaluru
Where our fellows work, research and publish
- SPAR
- AI Safety Camp
- MLCommons
- BlueDot Impact
- UNESCO
- ENAIS
- Pathfinder Fellowship
- AI Safety Atlas
- Sentient Futures
Find where to start
Two questions. We point you to the right first step.
New to AI safety? Read our AI safety career guide for India.
Your first stepA few hours a month
Speaker sessions and reading groups
Weekly reading groups, speaker sessions with guest researchers from frontier labs, and hackathons.
Weekly or biweekly3 local groupsYour first step5 to 7 hours a week, in STEM or building software
Technical AI Safety
AI capabilities are advancing faster than our ability to make them safe.
FreeBasic Python6 weeks, 5 to 7 hours a weekApplications open soonYour first step5 to 7 hours a week, in policy, law or economics
AI Governance
AI is being deployed faster than the rules governing it.
FreeNo coding6 weeks, 5 to 7 hours a weekApplications open soonYour first step5 to 7 hours a week, and not sure yet
A 1:1 career advising call
One-to-one guidance and next steps, before you pick a track.
One-to-one75+ sessions to date
Join an event near you
Live community sessions, offline regional hub meetups, and paper reading groups across India.
Research sprint · 15 November 2026
AI x Epistemics Research Sprint
Talk · 15 November 2026
AI Safety for Builders: Speaker Session with Ankur Pandey
Research sprint · 13 November 2026
AI x Epistemics Research Sprint
Research sprint · 24 October 2026
AI Collusion Research Sprint
Research sprint · 24 October 2026
AI Collusion Research Sprint
Moments from the ecosystem
Questions
Anything else? Write to contact@aisafetyindia.org.
What is AI Safety India (AISIN)?
How do I apply for courses and cohorts?
Are the courses and reading groups free?
Latest writing

AI Agents as Attack Vectors: Emerging Cybersecurity Risks for India
Understanding the risks of unauthorized AI-agent access and the misuse of agentic AI tools, based on CERT-In Advisory CIAD-2026-0049.

Find Your Way in AI Safety: Mapping Orgs, Frames, and Career Paths
Anthropic, DeepMind, MIRI, Redwood, GovAI. Each org sees "AI" differently. Each org sees "alignment" differently. That difference decides what job you'd do there. This session gives you a map. You will learn How...

Hackathon Winners Present: Do Multilingual Vision-Language Models Abstain under Cross-Modal Conflict in Low-Resource Languages?
What a hackathon-winning team found when they made VLMs choose between what they see and what they read.

AI Agents as Attack Vectors: Emerging Cybersecurity Risks for India
Understanding the risks of unauthorized AI-agent access and the misuse of agentic AI tools, based on CERT-In Advisory CIAD-2026-0049.

Find Your Way in AI Safety: Mapping Orgs, Frames, and Career Paths
Anthropic, DeepMind, MIRI, Redwood, GovAI. Each org sees "AI" differently. Each org sees "alignment" differently. That difference decides what job you'd do there. This session gives you a map. You will learn How...

Hackathon Winners Present: Do Multilingual Vision-Language Models Abstain under Cross-Modal Conflict in Low-Resource Languages?
What a hackathon-winning team found when they made VLMs choose between what they see and what they read.
Stay in the loop
Updates on cohort applications, research publications and fellowship deadlines.















