https://air-ml.org/research/

1<!doctype html><html lang=en-us><head><meta charset=utf-8><meta name=viewport content="width=device-width,initial-scale=1"><meta http-equiv=X-UA-Compatible content="IE=edge"><meta name=generator content="Hugo Blox Builder 5.9.7"><link rel=stylesheet href=/css/vendor-bundle.min.26c458e6907dc03073573976b7f4044e.css media=print onload='this.media="all"'><link rel=stylesheet href=https://cdn.jsdelivr.net/gh/jpswalsh/[email protected]/css/academicons.min.css integrity="sha512-IW0nhlW5MgNydsXJO40En2EoCkTTjZhI3yuODrZIc8cQ4h1XcF53PsqDHa09NqnkXuIe0Oiyyj171BqZFwISBw==" crossorigin=anonymous media=print onload='this.media="all"'><link rel=stylesheet href=https://cdn.jsdelivr.net/npm/[email protected]/dist/leaflet.min.css integrity crossorigin=anonymous media=print onload='this.media="all"'><link rel=stylesheet href=/css/wowchemy.f82ef6d8eb4a502b51e6f53c0c66c3ef.css><link rel=stylesheet href=/css/libs/chroma/github-light.min.css title=hl-light media=print onload='this.media="all"'><link rel=stylesheet href=/css/libs/chroma/dracula.min.css title=hl-dark media=print onload='this.media="all"' disabled><meta name=author content="Xiao Zhang"><meta name=description content="The mission of the AIR-ML Lab is to tackle the core research challenges in developing reliable and trustworthy AI systems through principled machine learning (ML) approaches, with a strong emphasis on resilience to adversarial attacks, interpretability, and robustness. Currently, the research of our lab is centered around Adversarial ML, focusing on three main directions:"><link rel=alternate hreflang=en-us href=http://air-ml.org/research/><link rel=canonical href=http://air-ml.org/research/><link rel=manifest href=/manifest.webmanifest><link rel=icon type=image/png href=/media/icon_hu1941449351133349273.png><link rel=apple-touch-icon type=image/png href=/media/icon_hu5505168350328901032.png><meta name=theme-color content="#1565c0"><meta property="twitter:card" content="summary"><meta property="twitter:site" content="@GetResearchDev"><meta property="twitter:creator" content="@GetResearchDev"><meta property="twitter:image" content="http://air-ml.org/media/logo_hu695335129701258550.png"><meta property="og:type" content="website"><meta property="og:site_name" content="AIR-ML"><meta property="og:url" content="http://air-ml.org/research/"><meta property="og:title" content="Our Mission | AIR-ML"><meta property="og:description" content="The mission of the AIR-ML Lab is to tackle the core research challenges in developing reliable and trustworthy AI systems through principled machine learning (ML) approaches, with a strong emphasis on resilience to adversarial attacks, interpretability, and robustness. Currently, the research of our lab is centered around Adversarial ML, focusing on three main directions:"><meta property="og:image" content="http://air-ml.org/media/logo_hu695335129701258550.png"><meta property="og:locale" content="en-us"><title>Our Mission | AIR-ML</title></head><body id=top data-spy=scroll data-offset=70 data-target=#TableOfContents class=page-wrapper data-wc-page-id=f1d044c0738ab9f19347f15c290a71a1>

1<script src=/js/wowchemy-init.min.4fef3e534144e9903491f0cc6527eccd.js></script>

1<aside class=search-modal id=search><div class=container><section class=search-header><div class="row no-gutters justify-content-between mb-3"><div class=col-6><h1>Search</h1></div><div class="col-6 col-search-close"><a class=js-search href=# aria-label=Close><i class="fas fa-times-circle text-muted" aria-hidden=true></i></a></div></div><div id=search-box><input name=q id=search-query placeholder=Search... autocapitalize=off autocomplete=off autocorrect=off spellcheck=false type=search class=form-control aria-label=Search...></div></section><section class=section-search-results><div id=search-hits></div></section></div></aside><div class="page-header header--fixed"><header><nav class="navbar navbar-expand-lg navbar-light compensate-for-scrollbar" id=navbar-main><div class=container-xl><div class="d-none d-lg-inline-flex align-items-center"><a class="navbar-brand d-flex align-items-center" href=/><img src=/media/logo_hu7186337307853674145.png alt=AIR-ML class=mr-2><span class=navbar-title>AIR-ML</span></a></div><button type=button class=navbar-toggler data-toggle=collapse data-target=#navbar-content aria-controls=navbar-content aria-expanded=false aria-label="Toggle navigation">
2<span><i class="fas fa-bars"></i></span></button><div class="navbar-brand-mobile-wrapper d-inline-flex d-lg-none"><a class=navbar-brand href=/><img src=/media/logo_hu7186337307853674145.png alt=AIR-ML></a></div><div class="navbar-collapse main-menu-item collapse justify-content-end" id=navbar-content><ul class="navbar-nav d-md-inline-flex"><li class=nav-item><a class=nav-link href=/#home><span>Home</span></a></li><li class=nav-item><a class=nav-link href=/research><span>Research</span></a></li><li class=nav-item><a class=nav-link href=/post><span>News</span></a></li><li class=nav-item><a class=nav-link href=/team><span>Team</span></a></li><li class=nav-item><a class=nav-link href=/project><span>Project</span></a></li><li class=nav-item><a class=nav-link href=/publication><span>Publication</span></a></li><li class=nav-item><a class=nav-link href=/contact><span>Contact</span></a></li></ul></div><ul class="nav-icons navbar-nav flex-row ml-auto d-flex pl-md-2"><li class=nav-item><a class="nav-link js-search" href=# aria-label=Search><i class="fas fa-search" aria-hidden=true></i></a></li></ul></div></nav></header><style>.navbar-title{font-size:1.8rem;font-weight:700;line-height:1.2}.navbar-nav .nav-link{font-size:1.1rem;font-weight:500}.navbar-nav .dropdown-menu .dropdown-item{font-size:1rem}.navbar-nav .nav-item{margin-left:.5rem}@media(max-width:992px){.navbar-title{font-size:1.6rem}.navbar-nav .nav-link{font-size:1rem}.navbar-nav .dropdown-menu .dropdown-item{font-size:.95rem}.navbar-nav .nav-item{margin-left:.6rem}}@media(max-width:768px){.navbar-title{font-size:1.4rem}.navbar-nav .nav-link{font-size:.95rem}.navbar-nav .dropdown-menu .dropdown-item{font-size:.9rem}.navbar-nav .nav-item{margin-left:.5rem}}</style></div><div class=page-body><article class=article><div class="article-container pt-3"><h1>Our Mission</h1><div class=article-metadata></div></div><div class=article-container><div class=article-style><p>The mission of the AIR-ML Lab is to tackle the core research challenges in developing reliable and trustworthy AI systems through principled machine learning (ML) approaches, with a strong emphasis on resilience to adversarial attacks, interpretability, and robustness. Currently, the research of our lab is centered around <span class=blue-highlight><strong>Adversarial ML</strong></span>, focusing on three main directions:</p><ul><li><p><span class=blue-highlight><strong>Foundations of Adversarial ML.</strong></span> Although rigor is quite essential for addressing trustworthy challenges in safety- and security-critical applications, the theoretical foundations of adversarial ML have lagged behind. The presence of adversaries often gives rise to non-i.i.d. learning problems, making the development of rigorous analytical frameworks far more challenging than in benign settings. Therefore, we aim to deepen the understanding of how ML algorithms behave in adversarial environments, leveraging insights from deep learning and probability theory. A primary focus is robust generalization in the presence of adversaries.</p></li><li><p><span class=blue-highlight><strong>Principled Algorithmic Design.</strong></span> A major component of our research is the design of principled a

2pproaches to rigorously audit the misbehaviors of machine learners and mitigate their vulnerabilities through the development of novel ML algorithms. Existing approaches are typically heuristic—insufficient to faithfully capture adversarial capabilities—and therefore remain unreliable against adaptive attacks. We aim to bridge this gap by designing theory-inspired auditing tools and defensive mechanisms that are robust to possible adaptive variations. Our current research focus is on generative modeling and optimization.</p></li><li><p><span class=blue-highlight><strong>Trustworthy AI Applications.</strong></span> We are dedicated to developing practical, high-performing ML tools to address critical trustworthy challenges across a variety of real-world AI applications, from computer vision (CV) and natural language processing (NLP) to biomedicine and cybersecurity. Key challenges include: (i) comprehensively yet faithfully characterizing adversarial behaviors within the context of specific applications, and (ii) designing trustworthy AI systems that are generalizable across diverse conditions, transparent in their decision-making, and resilient to adversaries (e.g., ensuring privacy and robustness), while preserving standard utility (e.g., clean performance, efficiency, and usability).</p></li></ul></div></div></article><style>.article-container{max-width:900px;margin:0 auto;padding:0 1rem}.article-style{font-size:1.05rem;line-height:1.6}.article-style p{margin-bottom:1em}.article-style ul{margin:0 0 1em 1.5em;padding-left:0}.article-style li{margin-bottom:.6em}.article-style p.bullet{padding-left:1.5em;text-indent:-1.5em;margin-bottom:.8em}.blue-highlight{color:#004f85!important}body.dark .blue-highlight{color:#80c1ff!important}</style><style>.article-style h1,.article-container h1{text-align:center;font-size:2.2rem;font-weight:500;margin-bottom:0;line-height:1.3}@media(max-width:768px){.article-style h1,.article-container h1{font-size:1.6rem}}</style></div><div class=page-footer><div class=container><footer class=site-footer><p class="powered-by copyright-license-text">Powered by Hugo Blox · Xiao Zhang © 2026</p></footer></div></div>

2<script src=/js/vendor-bundle.min.50933d940896e49f984a778650d5f7f5.js></script>

2<script src=https://cdn.jsdelivr.net/npm/[email protected]/dist/leaflet.min.js integrity crossorigin=anonymous></script>

2<script id=search-hit-fuse-template type=text/x-template>
3    <div class="search-hit" id="summary-{{key}}">
4      <div class="search-hit-content">
5        <div class="search-hit-name">
6          <a href="{{relpermalink}}">{{title}}</a>
7          <div class="article-metadata search-hit-type">{{type}}</div>
8          <p class="search-hit-description">{{snippet}}</p>
9        </div>
10      </div>
11    </div>
12  </script>

12<script src=https://cdn.jsdelivr.net/gh/krisk/[email protected]/dist/fuse.min.js integrity="sha512-o38bmzBGX+hD3JHWUFCDA09btWaqrNmoJ3RXLlrysA7PP01Kgs4UlE4MhelE1v5dJR3+cxlR4qQlotsW7jKsnw==" crossorigin=anonymous></script>

12<script src=https://cdn.jsdelivr.net/gh/julmot/[email protected]/dist/jquery.mark.min.js integrity="sha512-mhbv5DqBMgrWL+32MmsDOt/OAvqr/cHimk6B8y/bx/xS88MVkYGPiVv2ixKVrkywF2qHplNRUvFsAHUdxZ3Krg==" crossorigin=anonymous></script>

12<script id=page-data type=application/json>{"use_headroom":true}</script>

12<script src=/js/wowchemy-headroom.db4755770454eb63685f8de785c0a172.js type=module></script>

12<script src=/en/js/wowchemy.min.7f5ebaff62ae468cff8bb3dd1337bb9b.js></script>

12<script src=/js/wowchemy-map.a26e9d2f7238ba5b868384f1c5bc6477.js type=module></script>

12<div id=modal class="modal fade" role=dialog><div class=modal-dialog><div class=modal-content><div class=modal-header><h5 class=modal-title>Cite</h5><button type=button class=close data-dismiss=modal aria-label=Close>
13<span aria-hidden=true>&#215;</span></button></div><div class=modal-body><pre><code></code></pre></div><div class=modal-footer><a class="btn btn-outline-primary my-1 js-copy-cite" href=# target=_blank><i class="fas fa-copy"></i> Copy
14</a><a class="btn btn-outline-primary my-1 js-download-cite" href=# target=_blank><i class="fas fa-download"></i> Download</a><div id=modal-error></div></div></div></div></div>

14<script src=/js/wowchemy-publication.9c0e895144aef5a693008b5c5d450147.js type=module></script>

14</body></html>