May, 2026
Baarish spoke about Tattle’s manual evaluation methodology for evaluations of AI safety and the importance of combining human and automated evals for ensuring responsible deployments of AI in India’s social sector. This was an online panel curated by Digital Futures Lab during the AI and Equality Festival of Ideas event.
February, 2026
Tattle team met in Rishikesh to draft its AI Use Policy. Tarunima conducted some exercises to chart out the various operations that we engage in AI with the team.
February, 2026
At the AI Impact Summit, HumaneIntelligence and Tattle Civic Technologies hosted a panel discussion featuring Ashwani Sharma (Google.org), Sankat Verma (NumFOCUS Board of Directors), Mala Kumar and Tarunima Prabhakar on how open source software can lower barriers to red teaming
February, 2026
Tattle, Project Tech4Dev and The Agency Fund hosted a panel at the India AI Impact Summit 2026, discussing how to build cost-effective AI evaluation pipelines that maximize benefits and reduce harm in the development sector. Speakers: Baarish Aggarwal (Tattle), Vinod Rajasekharan (Tech4Dev), Edmund Korley (Agency Fund), Poorva Malviya (Noora Health), and Vineet Singh (Digital Green).
December, 2025
Kaustubha attended a strategy lab hosted by Equality Now in London on Algorithmic Harms and Misogyny to discuss legal and policy frameworks that could help address these harms.
December, 2025
Tattle conducted qualitative analysis on AI chatbot data from two grassroots NGOs to understand safety needs. Baarish, Denny and Kritika developed a framework for safety guardrails. Kritika and Denny built three kaapi-guardrails for slur detection, removal of personally identifiable information, and reduction of gender assumptions.
November, 2025
Siddharth Pillai, Kaustubha Kalidindi, and Tarunima Prabhakar co-authored a report on the impact of AI on ordinary women and gender minorities navigating the online world. Report was featured in the Guardian and cited by the Internet Freedom Foundation.
September, 2025
We are participating in UNDP’s AI Trust and Safety Reimagination Lab’s as a cohort member, engaging with multiple organisations around the world working to improve AI.
September, 2025
Tattle conducted a workshop on safety guardrails at Tech4Dev’s AI cohort.
August, 2025
Kaustubha attended DRAPAC 2025 in Malaysia and presented Tattle’s work on building AI safety datasets in Hindi.