Why Adopting AI in the Non-profit Sector Is Not a Choice Anymore?

Emerging technologies like AI have the potential to transform NGO operations, improve efficiency, and increase impact; NGOs that do not adopt AI risk reduced competitiveness, inefficiency, missed opportunities, and increased costs. 

Omdena’s CEO Rudradeb Mitra interviewed Janti Soeripto, CEO of Save the Children US, as part of the Omdena impACT leadership series. He asked about the differences she noticed while working in a Non-Governmental Organization (NGO) compared to a corporation (as she also worked at Unilever). Her answer should be an eye-opener for all NGO leaders:

Save the Children US CEO Janti Soeripto on the Omdena ImpACT Leadership Podcast

Six critical problems that keep NGOs from creating more impact

Lack of accountability

A lack of feedback mechanisms where users (communities and people in need) have no choice means there is no real drive for innovation. NGOs are prone to sticking with traditional approaches that have been used for many years. And this is something we at Omdena also observed while interacting with dozens of NGOs. 

In the words of Omdena’s CEO Rudradeb Mitra:

Funding 

NGOs often rely on grants and donations, which can limit their ability to invest in innovation and take risks. However, in recent years, we have observed an increasing availability of funding opportunities for data-driven solutions and projects based on machine learning techniques.

Several organizations provide funding and resources for AI projects for NGOs, including the Rockefeller Foundation’s Digital Impact Alliance program, the Open Society Foundations’ Information Program, the Chan Zuckerberg Initiative’s Justice and Opportunity Initiative, the Google AI Impact Challenge, the Microsoft AI for Good program, and the IBM AI for Social Good program. These initiatives provide technical assistance, grants, and other resources to help NGOs leverage the potential of AI and other digital technologies to address social and environmental challenges, promote open societies and human rights, and create a positive societal impact.

For example, Omdena has been chosen as the AI and Data partner of the consortium that won the USAID project to strengthen civil society and media capacities in Africa, LAC, Asia, MENA, and EE regions. 

Bureaucracy

Lengthy approval processes can slow the development and implementation of innovative ideas, while risk-averse decision-making can discourage experimentation and innovative thinking. 

However, also here we are witnessing an increasing number of positive examples. NGOs such as One Acre Fund, The Against Malaria Foundation, BRAC, and Médecins Sans Frontières have successfully reduced bureaucracy and improved efficiency by adopting innovative approaches such as mobile-based data collection, lean grant management, decentralized decision-making, and simplified organizational structures. These examples illustrate that NGOs can improve their operations by prioritizing efficiency, transparency, and local empowerment through innovative approaches.

Risk-aversion 

By not keeping up with technological advancements, NGOs risk falling behind their peers and losing relevance, as well as missing out on potential supporters and donors.

Especially the fear and inadequate comprehension of AI are major obstacles that hinder its wider adoption. The fear of AI is often exaggerated because it is based on fictional portrayals of AI in popular culture rather than an accurate understanding of what AI can and cannot do. In reality, AI is a tool that can be used to augment human capabilities and solve complex problems rather than a threat to humanity.

Feel free to check out hundreds of positive examples of how AI can be used in our project library. 

Lack of technical expertise

NGOs may not have the technical expertise or resources needed to implement innovative AI-based solutions.

However, NGOs and their staff can upskill themselves on AI through a variety of methods, including attending workshops, webinars, and training sessions focused on AI and machine learning. They can also participate in online courses and certification programs offered by universities and tech companies. Collaborating with AI experts and tech companies to develop and implement AI-based solutions can also provide valuable hands-on learning opportunities. Additionally, creating internal teams dedicated to exploring and implementing AI solutions can help build expertise and foster a culture of innovation within the organization. Finally, participating in open-source AI projects and networking with other NGOs can provide opportunities to share knowledge, best practices, and resources.

At Omdena, we follow a case study-driven approach to education using OmdenaAcademy and our local chapters’ approach of running open-source educational projects to address local challenges. 

Resistance to change 

NGOs may face resistance from stakeholders who are comfortable with existing practices or skeptical of new approaches. Resistance to change can be overcome when adopting new technologies by communicating the need for change, involving stakeholders in the decision-making process, providing training and support, and creating a culture of experimentation and learning.

How can NGOs use data and AI for greater impact?

Improving accountability and feedback mechanisms 

AI can help NGOs improve accountability and feedback mechanisms by providing real-time data analysis, automating monitoring and evaluation, and enabling predictive analytics, all of which can lead to more effective decision-making, better resource allocation, and improved program outcomes. For instance, AI-powered chatbots can facilitate communication with stakeholders and end-users (e.g., we built a Conversational AI Chatbot for People Affected by High Inflation and Increased Cost of Living). Machine learning algorithms can analyze large amounts of data and identify patterns and trends that can inform program design and implementation. This can help NGOs to continuously learn, adapt and improve their operations while ensuring accountability and transparency to their donors, beneficiaries, and partners.

A major issue is that data is often stored in papers, and there is almost no way to analyze and get insights from it. For example, we worked with a 96-year-old NGO to improve their case management system by transforming a large amount of paper-based data into digital records.

Understanding their users 

Data can provide information about what the users are thinking and feeling. We worked with Fondation Botnar to understand the aspirations of youth, where we engaged the youth to build the solution. 

In addition, NGOs can analyze social media data through sentiment analysis, social network analysis, natural language processing (NLP), and predictive analytics. This can provide valuable insights to inform their work and improve their impact. For example, we worked with NYU CIC on a hate speech detection project in Tamil Language using NLP.

In summary, AI can help NGOs to reach easily new users or those users who are not that easy to reach. We worked with the UNICEF Giga project to build an AI model to identify school locations via satellite images in South Sudan. It is rather impossible to find all school locations in a remote area. Once identified, the UNICEF Giga project plans to contact those schools to provide internet access.

Improving resource allocation

AI can be a valuable tool for NGOs in optimizing resource allocation. NGOs often operate with limited budgets and resources, and AI can help them make the most effective use of these resources. For example, AI can be used to predict areas of greatest need and allocate resources accordingly. It can also be used to monitor and track the impact of NGO programs, identifying areas of success and areas where resources may need to be adjusted. Additionally, AI can help NGOs analyze fundraising data and optimize their fundraising efforts to raise the necessary resources to meet their goals. Overall, AI can help NGOs improve their efficiency and effectiveness, allowing them to better serve their beneficiaries with limited resources.

Project examples: 

• Disaster Prediction to Optimize Relief Package Allocation
• Improving Access to WASH Services for Displaced and Vulnerable Communities
• Identifying and Visualizing Not-For-Profit Grant Data to Improve Matchmaking of Grants

Prediction and forecasting of emerging issues

By predicting and forecasting trends, needs, as well as disasters, NGOs can take proactive steps to address emerging issues.

Project examples: 

• Predicting Climate Change and Forced Displacement 
• Crop Yield Prediction using Machine Learning 

Measuring impact 

NGOs can use AI to measure their impact by analyzing data related to their programs and initiatives to identify trends and evaluate effectiveness. For example, an NGO working on poverty reduction could use AI to analyze data on poverty rates, employment rates, and education levels in a particular community to understand the impact of their programs and inform future initiatives.

For example, we have been working with Catholic Relief Services for more than 18 months in different project settings. In one project, the team developed a dashboard that uses NLP to track the conversation around getting children out of orphanages and into supportive homes and how COVID-19 is influencing the dialogue.

What process do NGOs need to follow to adopt AI?

What stops NGOs from investing in AI? 

Omdena has partnered with more than 70 NGOs, including prominent organizations such as UNICEF, UNHCR, Catholic Relief Services, and Save the Children. Throughout our collaboration, we have had the privilege of working alongside exceptional individuals, particularly those from the product or tech team, who are highly motivated and dedicated to innovation and progress. However, the implementation of technology on the ground and in all departments has been frustratingly slow. The reasons have been described earlier in the article, and that is why we developed a four-step process at Omdena to help civil society and development organizations get the most out of AI.

Omdena’s four-step process to bring AI to the real world

All phases are optional and customizable; however, many NGOs will benefit from going through several stages. 

In the Educate stage, Omdena provides education through courses, ideathons, workshops, and educational projects through Omdena School and Omdena Local Chapters. These educational initiatives are designed to equip organizations with the skills and knowledge necessary to tackle real-world AI challenges. This stage is fundamental to building a solid foundation for AI development. Organizations pose projects and research questions through our network of more than 130 Local Chapters in more than 60 countries to support capacity-building and community-driven action.  

In the Innovate stage, Omdena assembles diverse teams of 50 collaborators from all around the world to address real-world AI challenges through an eight-week AI Innovation Challenge. During this phase, teams work on prototyping AI models and may address research challenges through a Research Challenge. Omdena’s bottom-up approach ensures that each collaborator’s skills and experiences are leveraged to create innovative and effective AI solutions. The Innovate stage is where the real magic happens as these diverse teams come together to create impactful AI solutions.

The Deploy stage involves the top talent teams (top 2% of Omdena talent) that take the AI solutions that have been created and implement them for deployment. The collaborators might also utilize open-source AI and software solutions to ensure the solutions are scalable and accessible. Omdena’s best AI, data, and software engineers (chosen from 30,000 yearly applications and after 300 hours of Omdena projects). The Deploy stage is critical to ensuring that the AI models and solutions created during the Innovate stage can be effectively and efficiently deployed to address the real-world challenges they were designed for.

Finally, in the Scale stage, our partners can hire top performers from each project in-house, and the solutions are scaled to reach wider audiences. Omdena’s bottom-up approach ensures that every collaborator can contribute their unique perspective, and the most talented individuals from each project are retained to ensure maximum impact and scalability. The Scale stage ensures that the AI solutions created in the previous stages can be scaled to reach a wider audience and make a real impact.

It is not a choice anymore!

There are a lot of big problems that AI and machine learning can solve, which corporations and startups are not interested in or able to solve. A lot of hope lies in the NGO sector, and everyone working in the sector must realize that it is no longer a choice to use data, AI, or machine learning. One has to either innovate or slowly perish.