What does it look like when AI is done right?

Bidart: Respecting our traditions, our cultures, and our way of communication

Nicoll: Eliminating barriers to information

Shukla: Absolute synchrony between agriculture, apiculture, food security – all led by empowered women

Ravinutula: Highest quality primary care and health access in developing countries

Ruxin: A reflection of the people that are designing it and the information that it’s fed

AI is everywhere. In the news. In our social media algorithms. At home and work.  

But who has access to it? How is it being used? And for the benefit of whom?

We tackled those questions with five powerful social sector leaders from around the world at “Shaping the Future of Inclusive Growth,” a webinar to celebrate the awardees of the Artificial Intelligence to Accelerate Inclusion Challenge. The AI2AI Challenge is data.org’s fourth global innovation challenge, made possible with support from the Mastercard Center for Inclusive Growth. 

Here are three rules for designing and deploying AI with intention:

1. Lead with Local
   Leading with the local context is a pillar of our work at data.org, and was a clear priority in the AI2AI solutions. IDInsight depends on political buy-in and local expertise, scaling an AI-powered call center with more than 40,000 health extension workers providing real-time medical guidance on complex cases. BEEKIND, an initiative of Buzzworthy Ventures, includes humans in the loop for their AI app that troubleshoots beekeeping issues and optimizes hive placements.

2. Build Trust
   The International Rescue Committee interacts with users at the worst moment of their lives. Their AI tool, Signpost, provides lifesaving information to displaced people across  30 languages, supporting 400 trained people in providing personalized answers in plain language, and with recognizable branding specific to each of the 30 countries in which they operate. Link Health cuts through stigmatizing, overly complex, and duplicative benefit enrollment processes to increase access to critical federal aid benefits – $80 billion of which go unclaimed in the U.S. each year.
   
3. Think Outside the (Data) Box
   Strong AI tools rely on strong data. Eighty percent of businesses are informal or unbanked in Latin America, meaning they lack the information banks typically rely on to award loans. Women entrepreneurs are especially at risk of financial exclusion. Quipu set out to collect and analyze other meaningful inputs, like videos of the businesses.
   

Watch the webinar

These compelling takeaways from social impact leaders around the world reinforce why data.org seeks to source, support, and help scale innovative approaches that use data and AI to tackle some of the most intractable problems we face. Our global innovation challenges consistently identify groundbreaking solutions like these, making use of AI with a responsible lens and allowing us, as a platform for partnerships to democratize data, to begin to apply and grow those solutions to more people, in more places, across more sectors.