Aviral Srivastava is a Security Engineer at Amazon, specializing in offensive security research at the intersection of AI systems, exploitation, and applied machine learning. His work focuses on attacking AI - not hypothetically, but at the systems level, through vector collisions, LLM inference-time side channels, retrieval poisoning, token leakage, and large-scale misalignment strategies, etc. backed with theoretical cs
He holds a Master's degree in Cybersecurity Analytics and Operations from Penn State University, has co-authored/co-edited several books on AI security and red teaming, and has published over 30 peer-reviewed papers in top venues including ICLR, NeurIPS, and ACM. He was named an RSA Security Scholar (2025) and awarded Cybersecurity Innovator of the Year at BSides Bangalore.
Aviral builds tools for real red teams—such as:
ZeroDayForge – LLM-powered exploit and ROP chain generator
KernelGhost – kernel-mode eBPF rootkit and hypervisor escape lab
His speaking credentials include RSA Conference, CypherCon, CactusCon, and multiple BSides chapters, where he shares original research on breaking and defending AI systems.
This blog is for practitioners working at the sharp edge of AI and cybersecurity. Expect:
Novel offensive methodologies for attacking foundation models, inference layers, and retrieval stacks
Real vulnerability research, backed by systems traces, proofs of concept, and code
Elite-level exploitation techniques at the interface of ML, systems engineering, and cybersecurity
Research-grounded perspectives on adversarial robustness, alignment bypass, and AI-native threat models
No marketing. No AI news blabber. No watered-down content.
Every post is crafted for security researchers, ML engineers, red teamers, and technical architects who want to understand, replicate, and defend against real-world AI attacks.
🐙 GitHub: github.com/Aviral2642
💼 LinkedIn: https://www.linkedin.com/in/aviralsrivastava23/
