Chatbots popularity is partly due to their ease of use and increasingly “natural feel” thanks to advancements in machine learning and natural language processing. The chatbot market is expected to reach around
1.25 billion USD in 2025. After-sales and customer service bots are leading the charge.
But how journalists and fact-checkers look at the future of AI chatbots?
As an urgent response to the Turkey/Syria earthquake of February 2023, Fatabyyano, a fact-checking organization that serves fact-checks to an Arabic-speaking audience spanning across 19 countries, launched an
emergency tipline to deliver critical information to the earthquake victims. The automated messaging solution is powered by Meedan’s Check tool, which supports the development of AI-assisted chatbots. The collaboration marked the first Check tipline made for an Arabic-speaking audience. This emergency initiative has put in the foreground three pivotal questions that have the potential to inform future interventions: How fast can we deploy critical and efficient tiplines in times of crisis? How can we adapt to different contexts and audiences? How do we prioritize needs and join forces with other nascent initiatives?
In a tipline bot context, adapting mainly revolves around channels and language. Channels may vary depending on the country or region, as some populations use WhatsApp, while others use Viber. As for language, the tipline bot should be localized to engage with this specific audience. In some cases, multiple communities in the same country need access to fact-checked information in two or three languages.
This workshop will be a one-hour session in which attendees will examine existing chatbot structures and limitations, and examine what a ‘machine in the loop’ approach might look like for designing a chatbot within a newsroom. We will specifically design chat bots that can serve to inform communities in crisis contexts, examine unique needs in crisis settings, and prototype using Check.
The prototypes will be compared with conversations made with chatbots in different cultural and political contexts. If an informed citizen is the cornerstone of democracy, how can journalists make use of chatbots to deliver credible information and bring trust back to the conversation?
