Pitt Cyber Accelerator Grants (PCAG) provide initial funding for novel and innovative projects that advance Pitt Cyber’s mission: to bring the breadth of one of the world’s leading public research universities to bear on the critical questions of networks, data, and algorithms, with a focus on the ever-changing gaps among law, policy, and technology.
PCAG funding is to be used especially to advance projects that establish and extend key intellectual infrastructures, such as research and teaching in methods with broad application and utility across disciplines, and projects that bring together researchers, teachers, and/or students across fields of interest.
PCAG funding may be particularly useful in launching transdisciplinary projects that may later attract funding and sponsorship within the university, and/or from government, philanthropic, and industry partners.
Fall 2025 Awardees
Certified Security: Bridging Technology and Policy to Protect Critical Infrastructure
This project begins with the premise that certified software—applications mathematically verified to function as intended—can significantly reduce cyber risk, provided it is supported by the right policy frameworks. To advance this research and generate broader insights, we will host an interdisciplinary, problem-driven workshop where participants present concrete technical or policy challenges and collaborate on actionable solutions, producing a public report, scholarly dissemination, and building a cross-sector network on secure-by design systems.
Erica Owen, Associate Professor and Associate Dean, Graduate School of Public and International Affairs
Daniel Cole, Associate Professor, Swanson School of Engineering
Challenges of Generating Culturally Appropriate, Multimodal Persuasive Media
Given the surge of interest and the promising performance of text generation and image generation (T2I) models, their use for creating persuasive messages is imminent, but naive use would be problematic: (a) prompts to text-to-image methods are commonly visually concrete, with specific objects mentioned, yet a user seeking to generate a persuasive image may simply input an abstract message; (b) AI systems have well-documented biases, both in terms of propagating problematic stereotypes in Western culture, as well as not producing responses that are appropriate for use in other cultures; (c) persuasiveness, effectiveness, trust and potential harm may vary when viewers consume real vs AI-generated media. We will examine a diverse spectrum of persuasive media, from traditional ones like video/print advertisements, to modern social media, including multimodal tweets and influencer videos, and how they might be complemented by artificially generated persuasive content, in the context of these three challenges.
Adriana Kovashka, Associate Professor, School of Computing and Information
Diane Litman, Associate Dean for Mentoring and Development, School of Computing and Information
Christopher Maverick, Teaching Assistant Professor, Dietrich School of Arts & Sciences
Analyzing Harmful Language and Social Dynamics in Online Gaming Voice Chat
Online gaming is a popular space for entertainment and community, but is also known for hate speech, harassment and other harms which can affect young people and all players. This project will analyze social dynamics in a large dataset of voice chat between players in online gaming livestreams with a focus on identifying harmful language and toxicity. This analysis will inform interventions and design changes to counter the spread of hateful narratives and make online gaming more inclusive.
Michael Yoder, Teaching Assistant Professor, School of Computing and Information
Dan Villarreal, Assistant Professor, Department of Linguistics
Scott Kiesling, Professor, Department of Linguistics
Dmitriy Babichenko, Clinical Associate Professor, School of Computing and Information
Lorraine Li, Assistant Professor, School of Computing and Information
PittBusiness Cybersecurity Workshop: Transparency in the Boardroom
With an increasing number of cyber-attacks faced by businesses of all sizes, one of the largest threats continues to be cyber education, particularly in the boardroom and among executive leadership. Providing education and support from the top down is the best way to instill a culture of cyber awareness across an organization. This one-day conference will invite both local business leaders and PittBusiness students to an interactive day of education focused on improving cyber knowledge in the boardroom.
Brad Messner, Clinical Assistant Professor of Information Systems and Technology Management, School of Business
Leveraging LLMs and Rapport-Building to Counter Health Misinformation
This project explores how conversational AI can counter health misinformation while using rapport-building strategies to foster trust and understanding. It considers differences in health literacy and disparities in access to reliable information to ensure messages are clear and effective. The approach has the potential to improve public resilience against false health claims and support more informed health decisions across communities.
Jamie Zelazny, Assistant Professor, School of Nursing
Yu-Ru Lin, Professor, School of Computing and Information
Rr Nefriana, PhD Student, School of Computing and Information
Mitigating Activity Inference Attacks in Virtual Reality and Augmented
Virtual reality (VR) and Augmented Reality (AR) devices and applications are becoming increasingly prevalent. This project aims to address the privacy and security challenges for AR/VR users in multi-modal sensing environments. We will develop device interfaces and usable obfuscation techniques to protect against inference of user’s activities on VR/AR platforms by employing a customizable obfuscation model that can be configured to provide a desired tradeoff between privacy protection and usable VR/AR experience.
Balaji Palanisamy, Associate Professor, School of Computing and Information
Na Du, Assistant Professor, School of Computing and Information
Adam Lee, Professor and Vice Provost for Undergraduate Studies, School of Computing and Information