Founder · AI Safety Researcher
Aditya Raj

Building the pipeline the global AI safety field is missing — starting with India. Active researcher and field-builder, currently a SPAR Fellow working on technical AI safety research.

Current Research
SPAR Fellow — Ongoing
Capability Spillovers from AI Safety Research
A framework for analysing and forecasting how alignment research capabilities transfer to capabilities work.
ACL Workshop — Submitted
Inter-Agent Persona Exposure and Safety Alignment
Does exposing AI agents to each other's personas erode safety alignment? A controlled study on multi-agent Gemini systems.
Project Spillover
Quantifying the Alignment Tax
Controlled experiment showing naïve safety fine-tuning causes catastrophic capability collapse, validated via mechanistic interpretability.
Credentials & Experience
SPAR Research Fellow — Technical AI safety researcher Ongoing · International AI Safety Research Program
Jailbreak Hackathon — Top 30 globally Grayswan · Adversarial AI safety evaluation
Bluedot Impact — AI Safety Fundamentals, AI Governance, Biosecurity Three tracks completed · London-based AI safety accelerator
Co-operative AI Fellowship — Multiagent Systems & Coordination AI Safety Asia
Facilitator — AIxAnimal Fellowship, FutureKind Fellowship, AI Safety Collab Sentient Futures · Scaling Altruism
EAGx Volunteer — India 2024, Singapore 2025, India 2025 Effective Altruism Global conferences
Ran Cohort 1 — 30 researchers trained, 3 university clubs launched AI Safety India Community · 2025
Founder — EA NIT Agartala First Effective Altruism chapter at NIT Agartala
Open Position
Work on the
most important
problem. With us.

We are building the field-building infrastructure the global AI safety ecosystem is missing. If you want to be part of that — start here.

We reply within 12 hours.
Received. We'll be in touch within 12 hours.