Deepfake Detection in South Asia
Studying the technical and sociocultural gaps that limit existing deepfake-detection methods in South Asian settings, with an emphasis on culturally grounded datasets, benchmarks, and evaluation.
Machine Learning Engineer · Trustworthy AI Researcher
Hi, I'm Itba Malahat.
Building AI systems—and digital spaces—worthy of trust.
I build for Pakistan and beyond.
I'm a machine learning engineer and Research Assistant at the HCI Lab, University of Colombo School of Computing, working with Dr. Dilrukshi Gamage on deepfake detection, trustworthy AI, and media literacy for the Global South. I'm also part of the core team on the HCI Deepfake Detection for South Asia initiative.
I did my bachelor's in Computer Science at GIKI (Ghulam Ishaq Khan Institute of Engineering Sciences and Technology).
I like to learn and solve problems.
Research Interests
- Trustworthy AI
- AI Safety
- Multimodal Learning
- Deepfake Detection
- Computer Vision
- AI Governance
- Agentic AI
- Data Engineering
Selected Work
AVIA: Audio-Visual Integrity Analyzer
A unified deepfake-detection framework for images, videos, and audio, trained and evaluated across 12 benchmark datasets and more than 5,000 samples.
Deepfake Discourse Across Public Platforms
A mixed-methods study of how deepfake-related conversations, concerns, and harmful practices move from mainstream platforms into fragmented public channels.
Latest News
August 2026
Our paper, “Deepfake Detection in South Asia,” is in the final stages of write-up ahead of submission to ACM.
Our paper, “Deepfake Detection in South Asia,” is in the final stages of write-up ahead of submission to ACM.
July 2026
I began working with Dr. Subhayan Mukherjee, Dr. Dilrukshi Gamage, and the Data4Transparency (D4T) community on a mixed-methods study of deepfake discourse across Reddit and public Telegram channels, targeted for submission to ICWSM.
I began working with Dr. Subhayan Mukherjee, Dr. Dilrukshi Gamage, and the Data4Transparency (D4T) community on a mixed-methods study of deepfake discourse across Reddit and public Telegram channels, targeted for submission to ICWSM.
June 2026
I concluded my role as a Senior Technical Support Engineer at Motive after resolving more than 200 production incidents across cloud services, firmware, APIs, and AI-powered telematics.
I concluded my role as a Senior Technical Support Engineer at Motive after resolving more than 200 production incidents across cloud services, firmware, APIs, and AI-powered telematics.
December 2025
I joined Motive as a Senior Technical Support Engineer.
I joined Motive as a Senior Technical Support Engineer.
December 2025
Joined the Colombo HCI Lab as a Research Assistant under Dr. Dilrukshi Gamage, researching deepfake detection and media literacy for the Global South.
Joined the Colombo HCI Lab as a Research Assistant under Dr. Dilrukshi Gamage, researching deepfake detection and media literacy for the Global South.