About Rahul A R
Rahul A R is a practising advocate who pivoted into AI research during the pandemic, driven by the belief that technology should empower humanity. He is the founder of Dharmabot AI, an AI-powered legal assistance platform in India, and leads the Nova AI initiative at UB Intelligence — a long-term research program aimed at developing trustworthy artificial superintelligence using Small Language Models.
Education: BBA LLB (Hons) from Mar Gregorios College of Law (2017–2022); Diploma in Cyber Law from Asian School of Cyber Laws (2024).
Research & Insights
Red Teaming Advanced AI: An Empirical Vulnerability Analysis
Demonstrates that safety-aligned LLMs are not robust to the 13-protocol Jailbreak Framework. Found 87% of tested models failed at least 3 attack vectors including Contextual Exploitation, Role-Play Vulnerability, and Multi-turn Degradation.
Read full research
Nova AI System Development
Documents the Nova initiative — a unified Small Language Model family (Axion series) combining edge and cloud AI with agentic task orchestration, aimed at developing trustworthy ASI with constitutional AI embedded in pre-training.
Read full research
AI Psychosis and Chatbot Risks
Analysis of psychological risks of prolonged AI companion use: grandiose beliefs, emotional dependency, and reality distortion. Calls for Non-Deceptive AI standards and circuit-breaker design patterns.
Read full research
How AI is Empowering Cyber-Crime
Research on AI-automated spear phishing, deepfake-based biometric bypass, and polymorphic malware. Advocates for autonomous AI-driven Security Operations Centers as the primary defense.
Read full research