Tattle

Tattle is a civic technology organisation that builds open-source tools and datasets to understand and respond to misinformation and harmful content in India. It brings together technologists, researchers, and artists working on a healthier online information ecosystem, with a focus on content that circulates on chat apps and social media in Indian languages. All of Tattle's tools and datasets are openly available, and the organisation works collaboratively with fact-checkers, researchers, civil society groups, and journalists. Headquarters: India (distributed team) Sectors: Civic Technology, Misinformation Research, AI Safety, Media Literacy, Online Harms

What They Do Tattle's work spans three areas: building tools, creating open datasets, and conducting research on online harms.

Tools Uli is a browser plugin that lets users identify, flag, and respond to online gender-based violence on social media. It uses localised content moderation and is built for Indian languages and contexts. Users can redact slurs, archive abusive content, and find resources, giving individuals more control over their online experience without relying on platform action.

The Deepfakes Analysis Unit (DAU) is a collaborative platform for fact-checkers and media forensic experts to collectively assess possible deepfakes and AI-manipulated media. It supports a shared workflow for analysis and builds a labelled dataset of manipulated content over time.

Viral Spiral is an adaptive digital card game about identity, bias, and information sharing. It is designed to build media literacy skills by helping players understand how misinformation spreads and why people share it.

Kosh is Tattle's main archive of content scraped from fact-checking sites, social media, and chat apps including WhatsApp. It serves as a searchable database for researchers and fact-checkers tracking the spread of specific content.

Khoj is an image search tool that checks whether an image has previously appeared in a fact-checking article, helping identify recycled or misleadingly reused visuals.

Datasets Tattle builds and openly releases datasets relevant to online safety and misinformation in India, including a safety benchmark dataset in Hindi built as part of ML Commons' safety benchmark project. The DAU also produces a labelled dataset of deepfakes and media manipulation cases.

Research Tattle publishes research and blog posts on responsible AI, media literacy, online gender-based violence, and web scraping methods. Recent work includes an evaluation framework for AI tools in the social sector and research on online gender-based violence in India's gaming spaces.

Tattle is a good fit for people with backgrounds in software engineering, data science, NLP, research, design, or communications who want to work on civic technology and online safety in the Indian context.