PeopleofSuperintelligence

Researcher

Pan Zhou

Also known as 周潘

Pan Zhou is a professor at Huazhong University of Science and Technology (HUST) who studies the security of AI systems, from attacks on deep-learning models to LLM agents and robot-control models. He earned bachelor's and master's degrees at HUST, completed a PhD at Georgia Tech in 2011 under John Copeland, and worked at Oracle from 2011 to 2013. After joining HUST as an associate professor in 2013, he became a professor in 2020, and his papers include a 2018 study of service caching in mobile edge computing, a 2022 backdoor attack on crowd counting, and the 2025 BadVLA and ToolHijacker attacks. In 2026, he was invited to serve as an area chair for KDD 2026. As of October 2026, he directs HUST's Native Secure World Model Laboratory, which builds safety constraints into world models for embodied AI, autonomous driving, robot control and LLM agents.

Badges

Timeline

  1. 2002–2011

    Degrees at HUST and Georgia Tech

    He earns a bachelor's degree in communication and information systems in the advanced class at Huazhong University of Science and Technology (HUST) in 2006 and a master's degree in communications engineering there in 2008. He completes a PhD in electrical and computer engineering at the Georgia Institute of Technology in 2011 under John A. Copeland, with a thesis on power control and network capacity in cognitive radio networks. During the doctorate he works on foundational applications of Copeland's StealthWatch network-security system.

  2. 2011–2013

    Senior scientist at Oracle

    From 2011 to 2013, he works at Oracle Inc. in the United States as a senior technical scientist in the materialized view group.

  3. 2013–2020

    From associate to full professor

    He joins HUST in 2013 as an associate professor in the School of Electrical and Information Engineering, and moves to the School of Cyber Science and Engineering in 2019. In 2018, he coauthors "Joint Service Caching and Task Offloading for Mobile Edge Computing in Dense Networks" with Jie Xu and Lixing Chen for IEEE INFOCOM; the paper proposes OREO, an online algorithm built on Lyapunov optimization and Gibbs sampling. He becomes a professor in 2020.

  4. 2022

    Attacks on crowd counting and GNNs

    In 2022, he coauthors "Backdoor Attacks on Crowd Counting" with Yuhua Sun, Xingjun Ma, Lichao Sun and others, presented at ACM Multimedia; it proposes two density-manipulation backdoor attacks. His paper with Binghui Wang and Youqi Li on bandit-based black-box attacks against graph neural networks is accepted at CVPR 2022 as an oral presentation.

  5. 2024–2025

    A cybersecurity base in Wuhan

    From 2024 to 2025, he serves on temporary assignment as deputy director of the construction office of the National Cybersecurity Talent and Innovation Base in Wuhan.

  6. 2025

    BadVLA and ToolHijacker

    In 2025, he coauthors BadVLA with Xueyang Zhou, Lichao Sun and others, a backdoor attack on vision-language-action models that reaches near-100% attack success on several benchmarks with little loss of clean-task accuracy. With Neil Zhenqiang Gong, Lichao Sun and others, he also coauthors ToolHijacker, a prompt-injection attack on how LLM agents select tools, published at NDSS in 2026.

  7. 2026

    KDD area chair, AAAI award

    In January 2026, he is invited to serve as an area chair for KDD 2026. In 2026, AAAI-26 gives him an Outstanding Program Committee award.

  8. As of October 2026

    Directing a world-model lab

    He is a professor and doctoral supervisor in HUST's School of Cyber Science and Engineering and directs its Native Secure World Model Laboratory. The lab builds safety constraints into world modeling, so that AI systems for embodied intelligence, autonomous driving, robot control and LLM agents can predict how the environment will change and flag risks as they do.

Career

  1. Since 2013Huazhong University of Science and TechnologyAssociate Professor, Professor
  2. 2024–2025National Cybersecurity Talent and Innovation BaseDeputy Director
  3. 2011–2013OracleSenior Technical Scientist

Education

  1. 2011Georgia Institute of TechnologyPhD
  2. 2008Huazhong University of Science and TechnologyMaster's degree in communications engineering
  3. 2006Huazhong University of Science and TechnologyBachelor's degree in communication and information systems

From the biography.

  1. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Huazhong University of Science and Technology. We estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.
  2. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Huazhong University of Science and Technology. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  3. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Huazhong University of Science and Technology. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  4. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Georgia Institute of Technology. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Pan Zhou yet. Send us a free-licensed photo.

  1. 个人简历 - 原生安全世界模型实验室 nswm-lab.github.io. Checked 22 October 2026.
  2. 华中科技大学主页平台管理系统 Pan Zhou--英文主页--Home faculty.hust.edu.cn. Checked 22 October 2026.
  3. 华中科技大学主页平台管理系统 Pan Zhou--英文主页--Paper Publications faculty.hust.edu.cn. Checked 22 October 2026.
  4. [1801.05868] Joint Service Caching and Task Offloading for Mobile Edge Computing in Dense Networks arxiv.org. Checked 22 October 2026.
  5. 华中科技大学主页平台管理系统 Pan Zhou--英文主页--Paper Publications faculty.hust.edu.cn. Checked 22 October 2026.
  6. [2207.05641] Backdoor Attacks on Crowd Counting arxiv.org. Checked 22 October 2026.
  7. [2205.03546] Bandits for Structure Perturbation-based Black-box Attacks to Graph Neural Networks with Theoretical Guarantees arxiv.org. Checked 22 October 2026.
  8. [2505.16640] BadVLA: Towards Backdoor Attacks on Vision-Language-Action Models via Objective-Decoupled Optimization arxiv.org. Checked 22 October 2026.
  9. [2504.19793] Prompt Injection Attack to Tool Selection in LLM Agents arxiv.org. Checked 22 October 2026.
  10. 原生安全世界模型实验室 - 原生安全世界模型实验室 nswm-lab.github.io. Checked 22 October 2026.
  11. 华中科技大学主页平台管理系统 Pan Zhou--英文主页--Home faculty.hust.edu.cn. Checked 22 October 2026.