Mozilla Foundation

Mozilla Foundation is a US 501(c)(3) nonprofit founded on 15 July 2003 in Mountain View, California, as the institutional steward of the open-source Mozilla project after AOL's withdrawal from Netscape. As parent of the Mozilla Corporation — producer of Firefox — the Foundation operates simultaneously as a public-interest advocacy organization, an open-data infrastructure builder, and a fellowship and grant-funded capacity-builder for civil-society technologists. Since Nabiha Syed became Executive Director in May 2024 and the November 2024 restructuring refocused the organization, Mozilla has organized its public role around trustworthy AI as the frame for everything it does — applying what the Foundation calls the web-era infrastructure-plus-advocacy playbook to the emerging AI ecosystem.

Founding and structure

Mozilla Foundation was established as a California 501(c)(3) on 15 July 2003, receiving an initial $2 million from AOL alongside hardware and intellectual property to guarantee the independence of the open-source Mozilla project after AOL wound down its Netscape investment. The Foundation is the sole shareholder of the Mozilla Corporation — established 3 August 2005 to produce Firefox — and owns a five-part portfolio that now includes Mozilla Foundation, Mozilla Corporation, Mozilla Ventures (a $35M venture fund across 45 companies), Mozilla.ai (AI product development), and MZLA Technologies (Thunderbird). The organization describes this structure as a "double bottom line" model where commercial operations cross-subsidize mission-aligned work.

Mark Surman served as Executive Director from August 2008 and became Foundation President in 2022; Nabiha Syed was appointed Executive Director on 15 May 2024. In November 2024 the Foundation announced a roughly 30% reduction in its nonprofit workforce, with the advocacy and global programmes divisions restructured — a change Syed has framed as refocusing the organization on trustworthy AI and the open web. The Foundation's current board is chaired by Nicole Wong.

Trustworthy AI as organizing frame

Mozilla's Creating Trustworthy AI white paper, released in May 2020 after a year-long multi-stakeholder consultation, established the Foundation's theory of change on consumer AI: five guiding principles — agency, accountability, privacy, fairness, and safety — applied specifically to general-purpose internet products aimed at wide audiences. The paper's central argument — that "there is currently a harmful status quo in AI — one defined by profit before people, centralized power, and little transparency or accountability," in J. Bob Alotta's formulation — positions the Foundation as a movement actor rather than a neutral tech-policy voice. The Foundation's current strategy, articulated by Syed and Surman as "rewiring Mozilla to do for AI what we did for the web," centers on three pillars: Humanity First (protecting privacy and human agency), Open Collective Power (leveraging open-source and community models), and Alternative Economics (a double bottom line that advances mission while shaping markets).

AI literacy and public education

Mozilla's longest-running direct consumer-engagement program is Privacy Not Included, a product privacy guide launched in 2017 that gives consumers comparative privacy ratings of connected tech products, including AI-powered apps. The guide frames consumer choice as a lever for corporate accountability — providing technically accurate, even-handed information to general internet users to pressure companies to improve privacy practices. After the 2024 restructuring, Privacy Not Included migrated into product reviews at Nothing Personal, Mozilla's consumer tech magazine, where privacy experts continue to conduct hands-on testing. The Internet Health Report, published from 2017 onward, documents the state of web literacy, digital inclusion, privacy, decentralization, and openness — pivoting toward trustworthy AI as a focal theme from 2018 onward and serving as the Foundation's annual public-facing diagnostic of internet health.

Common Voice and open data infrastructure

Common Voice, launched in 2017, is Mozilla's most legible exercise of the infrastructure-as-advocacy approach: a crowdsourced multilingual speech corpus released into the public domain under CC0, built on the argument that opening up training data is one concrete way to level the playing field while mitigating AI bias. Common Voice has grown into the world's largest multilingual public-domain voice dataset, covering more than 100 languages including widely spoken and minoritised languages such as Welsh, Kabyle, and Sindhi. Under the 2024-era refocus, the Foundation is developing a data collective — a marketplace enabling controlled data sharing where dataset owners can be contacted for licensing — as the next iteration of its public-interest AI data infrastructure approach.

Fellowship and capacity-building programs

The Fellows for Tech & Society program supports up to ten fellows globally in twelve-month placements embedded with host organizations — universities, NGOs, advocacy groups, news organizations — to pursue projects on AI auditing, algorithmic accountability, open-source AI infrastructure, AI and democracy, content moderation governance, and bias and harm in machine learning. The 2024–2026 cohort is the current active group. The Democracy x AI Cohort — run through the Foundation's Incubator — provides $50,000+ per project for builders of AI systems that strengthen democratic practices: information quality, institutional transparency, civic participation, and protection of civic space. The 2026 cohort launched in June 2026 with a global field of applicants. The Responsible Computing Challenge is a higher-education curriculum-grant programme supporting faculty embedding social, political, and ethical context into computer-science teaching, expanded to South Africa in 2024 to seed responsible-computing capacity in the African higher-education system.

MozFest

Mozilla Festival (MozFest), begun in 2010 as the "Drumbeat Festival" in Barcelona before relocating to London in 2011, is the Foundation's annual global gathering for the open-web and internet-health community. The event has scaled to over 10,000 participants in its largest editions — the 2021 virtual format drew more than 10,000 participants across twelve days of community-led experiences — and organizes its programme around hands-on labs, forums, debates, showcases, and interactive installations across tracks aligned with Mozilla's internet health principles. The 2025 edition themed around "Unlearning" with tracks on harmful tech systems, power in tech governance, and traditional profit models; the 2026 edition returns to Barcelona (Recinte Fabra i Coats, 28–30 October 2026), framed as "the global AI policy summit, digital rights event, and open source festival for a healthier internet." MozFest's structural role in the movement is as a convening infrastructure that brings together activists, educators, artists, community organizers, journalists, and technologists from the global civil-society field in a format that explicitly runs alongside — and in tension with — industry-dominated AI conference circuits.

Policy advocacy

Mozilla's Open Policy & Advocacy team maintains an active policy blog and engages regulators across the US and EU. In February 2025, Mozilla served on the steering committee of the Paris AI Action Summit and co-launched ROOST (Robust Open Online Safety Tools) — open-source foundations for safer AI development — alongside Hugging Face, Microsoft, and OpenAI. In March 2025, Mozilla submitted comments to the White House AI RFI, calling for openness, accountability, and prevention of concentrated control by a few tech giants in AI development. On the EU AI Act, Mozilla has continued to push for proportionate open-source AI treatment and co-authored a guide for open-source AI developers navigating the Act's requirements with Hugging Face and the Linux Foundation. In digital markets policy, the Foundation filed comments urging faster UK DMCCA implementation — noting that Firefox daily active users on iOS rose 150% across the EU after the EU Digital Markets Act took effect — and supported California's Opt Me Out Act (AB 566) integrating Global Privacy Control. Mozilla ran a Digital Bootcamp with Cloudflare, Proton, and CENTR to educate European Parliament members on internet architecture, and submitted recommendations on the forthcoming Digital Fairness Act governing harmful design practices in AI-driven services.

Posture in the movement

Mozilla Foundation's distinctive contribution to the make-AI-good movement is that it operates as an advocacy actor, open-data infrastructure builder, and capacity-builder simultaneously from a single institutional frame. Unlike most digital-rights organizations, which either do advocacy or fund others, Mozilla builds and distributes things — Common Voice, Firefox, Privacy Not Included reviews, MozFest — at the same time as it lobbies, funds, and trains civil-society technologists. The Algorithmic Justice League is its most prominent funded civil-society partner on AI bias; Access Now is among the organizations it funds working on digital rights globally; the Electronic Frontier Foundation and European Digital Rights (EDRi) are its peers on digital rights policy. The Foundation's capacity to operate product-as-advocacy — using Firefox's market position to demonstrate that a privacy-first browser can scale, using Common Voice to show that open data can compete with corporate training corpora — is the theory of change that distinguishes it from every other actor in the make-AI-good landscape.