HerWILL has concluded its global AI for Digital Safety Datathon 2026 in Dhaka, naming Azra Tuni of Bangladesh as champion after a competition focused on using artificial intelligence to address harmful online content.
The programme brought together more than 2,100 participants from over 45 countries, according to organisers, with entrants asked to develop AI approaches capable of understanding complex online behaviour across Bangla, English and Banglish.
Nour Mohammad of Syria was named first runner-up, while Syed Zayan Anwar of Bangladesh took second runner-up. Organisers said the finalists reflected a broad group of emerging AI talent from Asia and Africa, with Bangladesh strongly represented among the top performers.
The datathon was designed to address gaps in existing content moderation systems, particularly in languages and contexts that are often underserved by global digital platforms. HerWILL said many moderation tools continue to struggle with non-Western linguistic environments, where meaning can depend heavily on context, code-switching and informal language use.
Participants took part in a structured programme that combined learning with practical application, working on real-world digital safety challenges. HerWILL said the initiative prioritised problem-solving, international collaboration and exposure to emerging AI frameworks. Most participants were students or early-career professionals.
HerWILL said the competition was part of a broader effort to expand access to AI capability beyond traditional centres of research and deployment. With most participants under the age of 25, organisers said the datathon highlighted a growing pipeline of young AI talent from regions that have historically been underrepresented in advanced technology development.
The organisation said its next phase would focus on turning global participation into long-term capability, with emphasis on responsible AI, multilingual digital environments and emerging international discussions on AI governance.
