February 14, 2026
Prompt and Circumstances: Evaluating the Efficacy of Human Prompt Inference in AI-Generated Art
This paper has been accepted at EvoMUSART 2026 and will be presented by Khoi Trinh.
Read moreSecretLab
University of Oklahoma
We develop secure, privacy-preserving, and trustworthy AI, software, and networked systems through rigorous, measurement-driven, human-centered cybersecurity research.
News
February 14, 2026
NinjaDoH Accepted at MADWeb 2026
NinjaDoH: A Censorship-Resistant Moving Target DoH Server Using Hyperscalers and IPNS has been accepted at MADWeb 2026 and will be presented by Scott Seidenberger in San Diego.
Read moreFebruary 13, 2026
SecretLab Website Launch
After waiting for years for AI to become more capable, we launched secretlab.page in one day using Codex-5.3, with modular PHP and JSON content for fast iteration.
Read moreProjects
Dataset Value Taxonomy
This project introduces a Dataset Value Taxonomy that rethinks how empirical research data are assessed in the age of AI, shifting emphasis from sheer dataset size to measures of temporal investment, scale, and accessibility to guide resource allocation. The framework helps researchers, funders, and evaluators prioritize and steward high-impact datasets by offering a calibrated scoring scheme that supports cross-field comparison while respecting discipline-specific norms.
LLM Package Hallucination Analysis
This work investigates software supply-chain risk created by hallucinated dependencies in code-generating LLM outputs. It combines large-scale measurement with actionable insights to reduce package-level attack surface in developer workflows.
Prompt Refinement for Image Regeneration
This research examines how people infer prompts from AI-generated art, how contextual circumstances influence inference quality, and how iterative human-driven prompt refinement affects regeneration outcomes. Through human-subject studies and controlled regeneration tasks, it identifies when prompt inference is reliable, where it fails, and how structured refinement workflows can improve quality, intent alignment, and user control.
EtherBee
EtherBee constructs a global Ethereum dataset that combines node performance measurements, honeypot interactions, and full network-session traces to support reproducible security analysis of P2P blockchain infrastructure. Building on this data, the project provides initial evidence of elevated reconnaissance activity against overlay-network nodes and characterizes attacker scanning behavior across the ecosystem.
Sponsors
Spiking Neural Networks in Federated Learning
This project studies the performance and efficiency trade-offs of spiking neural networks in vertical federated learning settings. It characterizes where SNN-based approaches are beneficial and where practical deployment constraints remain.
NinjaDoH
NinjaDoH develops a censorship-resistant moving-target DNS-over-HTTPS architecture using hyperscaler infrastructure and IPNS. The system is designed to improve resolver survivability and preserve access under active blocking pressure.
MagnetDB
MagnetDB provides longitudinal intelligence over the torrent ecosystem with structured metadata and IMDb-linked content signals. It enables longitudinal analyses of distribution patterns, ecosystem shifts, and content-network behavior.
Inbox Privacy Analytics
Inbox Privacy Analytics examines privacy implications of email marketing infrastructure used by online apps and services. The project quantifies tracking and exposure patterns to inform safer communication and data-handling practices.
VR Avatar De-anonymization
This project explores how non-VR motion side channels can undermine avatar anonymity in immersive environments. It highlights privacy risks in embodied interaction pipelines and motivates stronger identity-protection mechanisms.
AI Politicians
AI Politicians is a platform that builds fine-tuned language models to simulate political communication styles, policy positions, and debate behavior. The system combines parameter-efficient LoRA adaptation with retrieval-augmented generation to support factual, context-aware dialogue and structured one-on-one or moderated interactions.
People
Anindya Maiti
Director
Assistant Professor, Computer Science and Data Science
Khoi Trinh
Student
PhD, Data Science
Scott Seidenberger
Student
PhD, Data Science
Oluwasijibomi Ajisegiri
Student
PhD, Computer Science
Noah Pursell
Student
MS+BS, Computer Science
Kristiana Karshelieva
Student
MS+BS, Computer Science
Taylor Wise
Student
MS+BS, Computer Science
Roman Beames
Student
BS, Computer Science
Alee Lo
Student
BS, Computer Science
Frequent Collaborators
Murtuza Jadliwala
UT San Antonio
Associate Professor, Computer Science
Raveen Wijewickrama
UT San Antonio
Researcher, ScooterLab
Joe Spracklen
UT San Antonio
PhD Candidate, Computer Science
Greg Griffin
Oregon Department of Transportation
Principal Research Analyst
Bimal Viswanath
Virginia Tech
Associate Professor
Alumni
Clark Smith
Former Student
MS+BS, Computer Science
Narayana Phani Charan Nimmagadda
Former Student
MS+BS, Computer Science
Post Graduation: Paycom
Nima Najafian
Former Student
BS, Computer Science
Natalie Hill
Former Student
MS+BS, Computer Science
Post Graduation: Paycom
Preston Jones
Former Student
BS, Computer Science
Post Graduation: Blue Yonder
Koy Wilson
Former Student
MS+BS, Computer Science
Post Graduation: Paycom
Daylon Carse
Former Student
BS, Computer Science
Post Graduation: Heartland
Tristan Bony
Former Student
MS, Computer Science
Marc Beret
Former Student
MS, Computer Science
Post Graduation: Theodo Cloud
Luke Kirk
Former Student
BS, Computer Science
Post Graduation: Northrop Grumman
Fazil Raja
Former Student
MS+BS, Computer Science
Post Graduation: Paycom
Haidyn Peterson
Former Student
BS, Computer Science