Building for good: How civil society organizations are automating on Cloudflare
Captured source
source ↗Building for good: How civil society organizations are automating on Cloudflare | Cloudflare Blog
Skip to content
Tracking how governments target dissidents living in exile. Helping people in crisis find mental health support. Advocating for legislation that protects free expression online. These are a few examples of how some of the world's leading organizations are building the future of non-profit work with Cloudflare. AI is changing how people do their work. The goal of Cloudflare Impact is to help ensure that non-profit organizations are among the first to benefit. Today, we’re sharing what dozens of civil society organizations have built using our developer services with more than $7.5 million of Cloudflare credits. These stories show what is possible when AI applications are accessible, secure, and affordable to build and run. From "keep us secure" to "help us build" We believe a better Internet is one that allows people to express themselves online and access a diverse range of viewpoints. A key part of Cloudflare's mission has been making security services available for everyone and helping ensure that individuals and organizations working for the public interest are not forced offline by those more powerful. Today, Project Galileo , which provides free cybersecurity services to civil society organizations, protects more than 3,400 domains across more than 120 countries. Through these partnerships, organizations have shared with us how their needs have evolved from not only wanting to secure existing applications, but wanting to build new ones. AI has allowed non-technical teams to design, build, and scale tools tailored specifically for their workstreams and to advance their mission. This opportunity is arriving at a challenging moment for the sector. Many organizations report operating under financial strain after major reductions in government funding contributed to layoffs and closing of programs. As groups rebuild their work in a new environment, some have reported using AI to help do more with less. But adoption remains ad hoc. In a CIVICUS survey , more than half of civil society respondents viewed privacy concerns as a barrier to use. These groups hold sensitive data, like the location of activists or the identities of anonymous sources, and are disproportionately at risk of cyberattacks, according to our own Cloudflare data . Additionally, the cost of building and running complex automation workflows can be prohibitive. Among those surveyed by CIVICUS, 48% cited financial constraints limiting their adoption. Cloudflare helps address these concerns. Our developer services incorporate security and privacy protections from the outset. By running on our global network, applications are automatically protected from distributed denial-of-service attacks and attempts to gain unauthorized access to internal systems. Civil society groups can also build and run applications more cost effectively. Our lightweight and serverless architecture scales with demand without requiring organizations to provision or pay for idle infrastructure. Workers AI also removes the need to operate dedicated GPU infrastructure, while AI Gateway provides rate limits, observability into how teams are using AI, and intelligent model routing to reduce unnecessary model calls and keep costs under control. Awarding $7.5 million for the next generation of civil society work During Birthday Week last year, we expanded Cloudflare for Startups to include non-profit and public interest organizations, providing each of them with up to $250,000 in credits for our developer services. We are proud to announce that we selected 30 organizations to participate in our first non-profit, startup cohort, and we have been thrilled to watch their ideas and tools designed to tackle problems in climate science, humanitarian aid, mental health, civic engagement, and education come to life. Here is what a few of these non-profits have built. LebTown — local news, rebuilt on automation: LebTown is an independent non-profit newsroom covering Lebanon County, Pennsylvania. Until this year, the editorial team ran its entire operation across two Google Docs, a Google Calendar, Discord, and Gmail. LebTown is replacing that with a custom-built editorial management system that follows stories from pitch to publication, tracks audience reach, and lets reporters log community impact directly from their app or via Discord. The team has also built tools that convert county real-estate transfer records from PDF exports into structured data, and a real-time copyeditor agent that integrates with LebTown's WordPress site.
Kaya Guides — scaling mental health support: Kaya Guides built a WhatsApp-based mental health application that pairs people experiencing depression in India with trained lay counselors, making support accessible without the cost or waitlists of traditional therapy. As of September 2026, the application has supported 10,439 people, with 2,325 currently enrolled. To keep pace with demand, the team migrated its care management system to Cloudflare and now processes around 500,000 WhatsApp messages a month. They built a live AI system that listens to counseling calls and gives counselors real-time feedback to keep sessions on track. According to Kaya, these tools have helped develop a program that is more consistent and self-correcting as it scales, without significant additional cost or complexity. The Snorkelling Society — mapping the world’s snorkeling sites. The Snorkelling Society , is a UK non-profit building a web and mobile application for the global snorkeling community called SnorkelMap . This tool will help people discover new places to snorkel, explore information about different locations, and allow users to contribute their experiences. Because the application is built and maintained by volunteers, the Snorkeling Society uses Cloudflare to secure storage to allow for community contributions and uploaded images while also safeguarding against cyberattacks.
Working together to build automation tools for human rights We also heard from larger civil society organizations who wanted to build complex workflows and welcomed extra engineering support. These projects included several variations of the same problem: large amounts of manually-collected, dispersed information that needed to be integrated, reviewed, and structured in order for effective analysis to be conducted. The data involved was...
Excerpt shown — open the source for the full document.
Notability
notability 2.0/10Corporate blog post, not AI research or model release.