Please join Charlottesville Data Science on the evening of Wednesday, November 11 for the talk How AI Systems Leak the Data They’re Built On from Hunter Moore, co-founder and CTO of Hardshell, a Charlottesville-based AI security startup. In his talk, Hunter will walk through how deployed AI systems are attacked in practice, using real-world case studies to explore indirect prompt injection, system prompt extraction, and privacy attacks that induce a model to divulge sensitive data.
We’ll be gathering in person at Vault Virginia. We look forward to seeing you there!
About the talk
Retrieval-augmented generation made it easy to point a language model at an organization’s documents. It also made every one of those documents a potential instruction. Sure, the model was trained to refuse harmful requests, but what happens when the attack arrives inside an email, a web page, or a PDF the system retrieves on its own?
This talk walks through how deployed AI systems are attacked in practice – indirect prompt injection, data exfiltration through rendered output, system prompt extraction, and privacy attacks that pull sensitive data back out of a model. Each one is grounded in a real case, from hidden web text that hijacked Bing Chat to the ruling that held Air Canada responsible for what its chatbot told a customer. We will close with which defenses hold up, which ones break under an adaptive attacker, and what a data science team can do about it right away.
No security background required. Familiarity with LLMs and RAG will help.
About the speaker
Dr. Hunter Moore is the founder and CTO of Hardshell. He holds a Ph.D. in Systems Engineering from the University of Virginia, with expertise spanning AI security, machine learning, and autonomous systems.
He has over five years of experience advising the U.S. Department of Defense on AI security, including contributions to foundational policy and protocol development at the Joint AI Center and the Chief Digital and AI Office. His work has encompassed domains such as trustworthy and robust AI, generative AI, predictive and explanatory modeling, reinforcement learning, human systems, and systems engineering of autonomous systems.
He has taught AI-related courses for the U.S. Air Force, U.S. Army Special Operations Command (USASOC), and the University of Virginia, where he also designed and delivers a graduate-level AI security course and supervises capstone teams on trustworthy AI and AI security related topics.
Call for Demos: Show Us Your Agent Skills
Are you using AI agents in your data-centric workflows? If so, how? What’s working, and what’s not? Consider giving a quick demo of your agentic AI skills at a future Charlottesville Data Science event!
One of my favorite data science podcasts is Vanishing Gradients by Hugo Bowne-Anderson. This year, Hugo has been running an excellent series called Show Us Your (Agent) Skills, in which long-time Python, ML, AI, and data builders show how they’re actually using AI agents today. I’d love to replicate this idea within the Charlottesville Data Science community.
If you’re using AI agents to assist with data analysis, machine learning development, data engineering, or any other data-centric workflow, and you’re game to give a brief demo for your fellow community members in a friendly environment, we’d love to hear from you. Please submit your idea through the Charlottesville Data Science Call for Proposals!
More upcoming events around Charlottesville
Wednesday, October 28: Cville AI Explorers hosts Ain’t Misbehavin’: Keeping AI Agents Out of Trouble at Studio IX. Ben Cutter of Hardshell and Carson Sweet will compare training an agent to behave safely with putting technical controls around what it can do. They’ll cover guardrails and practical sandboxing for local agents, and test both on a deliberately troublesome agent.
Friday, October 30: University of Chicago professor Chenhao Tan, who directs the Chicago Human+AI Lab, gives the colloquium Science in the Age of AI at the UVA School of Data Science. He’ll look at what happens to peer review as AI-generated research grows, at the capabilities and limits of autonomous “AI scientists,” and at how the scientific ecosystem may need to change to realize AI’s promise for discovery.
Thursday, November 5: UVA’s digital humanities community hosts Princeton’s Meredith Martin, director of Princeton’s Center for Digital Humanities, for the free public lecture Reading Poetry’s Data in Bryan Hall. Drawing on her book Poetry’s Data, Martin argues that working carefully with large, messy humanities corpora is good training for using AI in research. She’ll make the case for knowing what you’re searching in as well as what you’re searching for.
Wednesday, November 11: Charlottesville Data Science hosts our November event at Vault Virginia. Hunter Moore, co-founder of Hardshell, a Charlottesville-based AI startup, will present the talk How AI Systems Leak the Data They’re Built On. The presentation will explore how motivated users can easily induce AI systems to disclose sensitive information from their training data and retrieval corpus.
Friday, November 13: UVA Data Science hosts Datapalooza 2026, their signature annual event showcasing research and industry trends across Grounds and the greater data science community. This year’s Datapalooza will focus on what the future of work looks like when data, AI, and human expertise come together.
Tuesday, November 17: rvatech/ holds its annual rvatech/Women in Technology Conference at the Science Museum of Virginia in Richmond.
Wednesday–Thursday, November 18–19: UVA’s Contemplative Sciences Center hosts AI Anxiety + Human Flourishing, a two-day public symposium on preparing young adults for a future with AI. Speakers include Reddit co-founder Alexis Ohanian, AI ethics researcher David Danks, and UVA President Scott Beardsley.
Tuesday, December 1: The Tom Tom Foundation hosts The Infrastructure of Tomorrow: Data Centers, AI, and Virginia’s Future, a conversation with State Senator Creigh Deeds and Delegate Katrina Callsen about whether Virginia can remain a leader in the digital infrastructure economy — while also protecting affordability, sustainability, and the long-term interests of its communities.
Friday, December 4: The LaCross Institute for Ethical Artificial Intelligence in Business at UVA Darden hosts its fourth annual UVA Conference on Ethical AI in Business at The Forum Hotel, bringing together leaders from business, technology, and policy to talk about adopting AI responsibly. This year’s theme is “The AI Readiness Imperative: Preparing People & Organizations for Responsible AI.”