Event Recap: “Innovative Uses of AI by Civil Society in Europe”

On June 26, GLOBSEC hosted an online discussion highlighting the innovative uses of AI by civil society organizations in Europe, exploring tools and technologies from leading tech companies designed to support these initiatives, and addressing the ethical challenges and concerns associated with AI in civil society. If you missed it or want to revisit the key points, read through the event summary below, which includes both a recap and a list of specific AI tools discussed. 

The session opened with moderator Marc Faddoul from AI Forensics, who underscored the ubiquitous role of AI in both our personal and professional spheres. He pointed out the need to consider not only the misuse of AI for disinformation and other harmful activities but also its potential to assist individuals and organizations. 

Kateřina Anna Magnna from Microsoft introduced the latest AI tools and platforms developed by Microsoft to aid civil society initiatives. She stressed the importance of mastering these tools to enhance efficiency, noting that tasks that once took an hour can now be completed in seconds with AI. She noted that this technology is beneficial for organizations of all sizes, enabling them to automate routine tasks, generate actionable insights from data, streamline decision-making, engage more effectively with stakeholders, and enhance security measures. 

Microsoft also recognizes the resource constraints faced by NGOs, particularly those with limited budgets and staffing. To address this, they have rolled out a series of trainings, tools, and hubs to help NGOs effectively utilize AI. All of these are listed below in detail. Kateřina also mentioned that Microsoft has special rates for NGOs and reduced prices and licenses. She noted too that in case of questions or suggestions, NGOs can contact Microsoft directly either via its “Nonprofit Community” or philanthropies team. 

Ognjan Denkovski from Democracy Reporting International (DRI) shared how DRI applies AI in their work, providing a practical example that could be replicated by other civil society organizations. He discussed their use of AI tools for topic modeling, analyzing the toxicity of online content, machine translation, and early detection of technologies that could be used for disinformation. All are detailed further in the list below. 

Ognjan also explored DRI's recent monitoring of the European Parliament elections, applying topic modeling and toxicity analysis across eight countries and generating an aggregated toxicity score. Their findings, available online, highlight the prevalence of toxic language in the discourse. 

He also referenced a less recent project, Disinfo Radar, using AI tools developed in-house to assess the potential impacts of emerging tech tools on disinformation at an early stage. The toolset developed automates the collection and analysis of data from electronic preprint repositories (arXiv), industry papers (syncedreview.com), and policy publications (IEEE), creating clusters of tech-related topics and calculating disinformation-potential scores based on dataset mentions. 

Paola Gálvez Callirgos, a Global Independent Consultant in Tech Policy & AI Governance, shared insights on leveraging AI tools to enhance CSO operations. She outlined strategies for using AI to enhance fundraising efforts, boost creative capacities, clearly communicate program impacts, and strengthen community connections. 

Paola highlighted a key issue in fundraising: its highly transactional nature, which often discourages repeat donations. This has led to a reactionary approach in CSO operations, where the urgency in communication contributes to high burnout rates among development staff while donor retention remains low. This dynamic, she noted, is damaging the sector. However, Paola stressed that AI holds tremendous potential to revolutionize fundraising, creating a more meaningful interaction for every stakeholder involved and ensuring the sustainability of the relationships that organizations value most. Tools like Fundraising.ai are pivotal, as they can predict who is likely to donate within the next 12 months and identify donors that might not be suitable for contact. 

Paola also pointed out the critical role of data and the ability to analyze it swiftly. Many NGOs lack dedicated data analysts, making AI an invaluable asset as it enables organizations to generate summaries and analyses.  

List of useful AI tools mentioned during the discussion: 

• Digital Skills Resource Hub - A resource for AI, cybersecurity, green skills, and learning materials. 
   
• TechSoup: Microsoft Digital Skills Center for Non-Profits - Webinars, recordings, guides, and best practices on Microsoft AI tools. 
   
• LinkedIn Learning Certificates and Learning Paths - Courses in AI, cybersecurity, and other digital skills. 
   
• DemoDays: Microsoft Tools and Technologies for Nonprofits - Live and recorded webinars on Microsoft's AI tools. 
   
• LinkedIn Nonprofits Resource Hub - Guides and best practices for maximizing LinkedIn's potential. 
   
• AI Explained - A resource to learn the fundamentals of generative AI. 
   
• TechSoup: Exploring AI with Microsoft Tools - Discounted learning paths for nonprofits. 
   
• Nonprofit Community - A discussion forum exclusive for nonprofits, offering best practices, events, and networking opportunities. 
   
• AI Trainer Toolkit - A toolkit for trainers to empower others with AI skills. 
   
• AI Learning Hub - Curated training resources tailored for professional roles or specific technologies. 
   
• AI for Educators - A resource to understand the basics of generative AI. 
   
• AI Compass for Nonprofits - A tool to measure the digital maturity level of organizations. 
•  

• Nonprofit Training and Resources - Self-paced and on-demand trainings on Microsoft solutions.
   

• BERTopic - A tool for topic modeling that integrates with OpenAI and Langchain for generating topic representations. 
   

• Perspective API - Part of Google's Jigsaw initiative, this tool conducts toxicity analysis in nearly 20 languages. 
   

• OPUS MT Models - Machine translation tools based on the transformer architecture, useful for many language pairs but relatively slow. 
   

• Fundraising.ai - Predicts potential donors likely to donate within the next year and identifies those NGOs should avoid contacting.