Nicholas Carlini

Nicholas Carlini is a preeminent researcher in the fields of computer security, machine learning, and privacy, whose work has fundamentally reshaped the understanding of vulnerabilities in machine learning systems. His groundbreaking research has advanced the study of adversarial attacks, data poisoning, model extraction, and the broader intersection of AI and security. Currently, he is a Research Scientist at Google Brain, where he leads efforts to develop secure and robust AI systems that can withstand real-world threats.

Nicholas's research is characterized by its depth, rigor, and practical impact. He has uncovered critical vulnerabilities in widely-used AI systems, such as adversarial examples—subtle perturbations to input data that can cause machine learning models to make incorrect predictions. His work on data poisoning has demonstrated how attackers can manipulate training data to compromise model performance, while his research on model extraction has revealed how adversaries can steal proprietary machine learning models by querying them. To counter these threats, Nicholas has developed innovative defenses, including robust training algorithms, anomaly detection techniques, and cryptographic methods for securing machine learning pipelines.

Beyond his technical contributions, Nicholas is a vocal advocate for responsible AI development. He has spoken at numerous conferences, including NeurIPS, ICML, and Black Hat, where he emphasizes the ethical implications of AI and the need for systems that prioritize safety, fairness, and transparency. His work has been featured in major media outlets such as The New York Times, Wired, and MIT Technology Review, and has influenced both academic research and industry practices. His insights have helped shape policies and guidelines for deploying AI in sensitive domains, such as healthcare, finance, and autonomous systems.

Nicholas is deeply engaged with the research community. He serves on the program committees of top-tier conferences, including NeurIPS, ICML, and IEEE S&P, where he helps shape the direction of research in AI security. He has also mentored numerous students and early-career researchers, many of whom have gone on to make significant contributions to the field. His dedication to advancing AI security has earned him widespread recognition, including prestigious awards such as the USENIX Security Best Paper Award and the IEEE S&P Distinguished Paper Award.

In addition to his research, Nicholas is an accomplished educator. He has taught courses on machine learning security at leading universities and has developed open-source tools and resources to help practitioners secure their AI systems. His commitment to education and outreach reflects his belief that the security of AI systems is a shared responsibility that requires collaboration across academia, industry, and government.