Social Insight Lab

Golisano College of Computing and Information Sciences | Rochester Institute of Technology

PI: Dr. Ashiqur R. KhudaBukhsh | axkvse@rit.edu | Google Scholar

About the Lab

Our lab lies at the intersection of artificial intelligence, machine learning, and natural language processing (NLP). We focus on evaluating and advancing the reliability, safety, and security of modern AI systems. Our work develops auditing frameworks and empirical methods to address real-world application challenges across public policy, cybersecurity, business, social sciences, and online ecosystem safety. Apart from publishing in top AI and NLP venues (AAAI, IJCAI, ACL, EMNLP, WWW) and broad interest journals (PNAS, IJOC, Patterns), our lab's research has been covered by more than 500 international media outlets including coverage from CNN, BBC, ABC News, The New York Times, Wired, to list a few.

Core Research Pillars

LLM Safety & Bias Auditing

Developing robust frameworks to stress-test large language models for implicit bias, racism, misogyny, and antisemitism, ensuring safer generative deployment.

NLP for Social Impact

Surfacing "hope speech" and positive dialogue in hostile digital spaces, quantifying societal polarization, and modeling media stance on critical public interest topics.

Data & Model Reliability

Investigating annotator variance, auditing offensive speech classifiers, and handling noisy user-generated data across low-resource multilingual context.

Selected Publications

Measuring Criticism of the Police in Local News Media Using Large Language Models
Logan Crowl, Sujan Dutta, Ashiqur R. KhudaBukhsh, Edson Severnini, Daniel S. Nagin
PNAS: Proceedings of the National Academy of Sciences (2025)
Impact Factor: 10.8
What About the Scene With the Hitler Reference? HAUNT: A Framework to Probe LLMs’ Self-consistency in Closed Domains Via Adversarial Nudge
Arka Dutta, Sujan Dutta, Rijul Magu, Soumyajit Datta, Munmun De Choudhury, Ashiqur R. KhudaBukhsh
ACL 2026
Featured in Conversation, UXXI News NPR
How Can You Tell if Your Large Language Model Could Be a Closet Antisemite? An Explainability-Based Audit Framework for Implicit Bias
Arka Dutta, Reza Fayyazi, Shanchieh J. Yang, Ashiqur R. KhudaBukhsh
AAAI 2026 (AI for Social Impact Track) Oral Spotlight (7.5%)
ARTICLE: Annotator Reliability Through In-Context Learning
Sujan Dutta, Deepak Pandita, Tharindu Cyril Weerasooriya, Marcos Zampieri, Christopher M. Homan, Ashiqur R. KhudaBukhsh
AAAI 2025
Navigating the Rabbit Hole: Emergent Biases in LLM-Generated Attack Narratives Targeting Mental Health Groups
Rijul Magu, Arka Dutta, Sean Kim, Ashiqur R. KhudaBukhsh, Munmun De Choudhury
COLM 2025
Featured in Mad in America
Down the Toxicity Rabbit Hole: A Framework to Bias Audit Large Language Models
Arka Dutta*, Adel Khorramrouz*, Sujan Dutta, Ashiqur R. KhudaBukhsh
IJCAI 2024 (AI for Good Track) Oral (20%)
Featured in CNN, ABC News, Wired, Good Morning America
Vicarious Offense and Noise Audit of Offensive Speech Classifiers: Unifying Human and Machine Disagreement on What is Offensive
Tharindu Cyril Weerasooriya*, Sujan Dutta*, Tharindu Ranasinghe, Marcos Zampieri, Christopher M. Homan, Ashiqur R. KhudaBukhsh
EMNLP 2023
‘Beach’ to ‘Bitch’: Inadvertent Unsafe Transcription of Kids’ Content on YouTube
Krithika Ramesh, Ashiqur R. KhudaBukhsh, Sumeet Kumar
AAAI 2022 (AI for Social Impact Track)
Featured in Wired, Indian Express
We Don't Speak the Same Language: Interpreting Polarization Through Machine Translation
Ashiqur R. KhudaBukhsh*, Rupak Sarkar*, Mark S. Kamlet, Tom M. Mitchell
AAAI 2021 (AI for Social Impact Track)
Featured in The New York Times, Wired
Voice for the Voiceless: Active Sampling to Detect Comments Supporting the Rohingyas
Shriphani Palakodety*, Ashiqur R. KhudaBukhsh*, Jaime G. Carbonell
AAAI 2020 (AI for Social Impact Track) Oral Spotlight (5.74%)
Featured in VentureBeat, Digital Trends

Active Grants & Sponsored Research

Trustworthy AI for Social Good (REU Site)

Funder: National Science Foundation (NSF) | Duration: 2025–2027

Role: Senior Personnel (PI: Daqing Hou) | Award: $464,781

Auditing Large Language Models for HARM and Mitigating Strategies

Funder: Lenovo | Duration: 2025–2026

Role: Principal Investigator | Award: $75,000

AI-Driven Detection and Tracking of Damaged Items During Relocation

Funder: BizCloud-RIT PRAXIS | Duration: 2025–2026

Role: Principal Investigator | Award: $40,000

Lab Members & Mentorship

Sujan Dutta

Role: Ph.D. Researcher (RIT)

Publications in IJCAI, EMNLP, AAAI, ACL, and PNAS.

Arka Dutta

Role: Ph.D. Researcher (RIT)

Publications in AAAI, IJCAI, ACL, EMNLP, ICWSM, and COLM.

Soumyajit Datta

Role: Student Researcher (RIT)

Publications in AAAI, WWW, EMNLP, and ACL.

Mallikarjuna Tupakula

Role: Student Researcher (RIT)

Publications in ASONAM and IJOC.

Contact

Department: Department of Software Engineering (Data Science Division)

Institute: ESL Global Cybersecurity Institute & GCCIS, Rochester Institute of Technology

Email: axkvse [at] rit.edu