Mozilla’s new AI strategy marks a return to its ‘rebel alliance’ roots

As Big Tech races to weave AI into nearly every product, Mozilla is betting some users want the opposite: the ability to turn it off.

Last week, the company announced new controls to allow users of its Firefox browser to decide when to use AI. When Firefox 148 debuts later this month, users will be able to manage or disable individual AI features like translations, tab grouping and a sidebar for chatbot like Claude, ChatGPT, Copilot, Gemini and Le Chat Mistral.

Much of Mozilla’s vision around AI was outlined in its annual State of Mozilla report, which was released last month and calls for a new Star Wars-style “rebel alliance” composed of developers, cybersecurity experts, investors, and others focused on responsible tech. The plan involves doing for AI what Mozilla once did in the earlier days of the web.

The goal is to “bend history in a different direction with the resources and the community we have,” says Mozilla Foundation president Mark Surman. In a recent interview with Fast Company about the strategy, Surman likened the winner-takes-all mindset of some AI giants and startups to the galactic empire’s ambition to have an expanding footprint. 

“The Empire, like any empire, is more diffuse and more spread out than you think it is,” Surman says. “Transforming things is a constant battle of trying to do stuff that’s for humanity, against the things that are threatening us and holding us back.”

Funding the rebellion

With more than 200 million users, Firefox is now Mozilla’s most popular product. However, Mozilla’s portfolio also includes other aspects like an email platform, a VPN, an AI data exchange, a venture arm and other initiatives for open-source AI. Mozilla also recently announced a new program inviting technologists to apply for a few months of paid work exploring early-stage ideas that could be worth Mozilla investing in.

Part of Mozilla’s plan includes spending around $650 million this year, with 80% going to improve and maintain core products like Firefox and the rest directed toward what Surman calls “systematic and aggressive” investments in trustworthy AI and related areas. Mozilla also has $1.4 billion in reserves that it could use as “dry powder” for worthy bets on things like open-source AI developer tools and encrypted AI assistants. But that’s not much compared with the hundreds of billions Mozilla’s rivals invest in AI-related capital expenditures each year.

While Mozilla has leaned on Star Wars’ “rebel alliance” metaphor before, its vision has roots in an era that now feels a long time ago (and far, far away). In 1998, when Netscape created Mozilla.org, Microsoft was on trial for antitrust, as early open-source projects began challenging proprietary control of the web. Surman recalls it feeling impossible at the time to unseat a company that dominated browsers, servers, and operating systems. (A few years after AOL bought Netscape, Mozilla was spun off in 2003 as an independent nonprofit, followed in 2005 with the creation of Mozilla Corporation as a for-profit subsidiary.)

“[It took] a set of people who all wanted a different future they could configure and tweak and make their own,” Surman says. “It’s not like they all had to build one big thing. We built a browser. A bunch of people built Linux, a bunch of people built web servers, and people built thousands of other things.”

Decades later, it’s now Google that’s on trial for antitrust while Mozilla competes against other privacy focused browsers like DuckDuckGo and Brave alongside AI startups like OpenAI and Perplexity that now have their own browsers. The antitrust scrutiny and growing distrust of AI and Big Tech have some finding a new hope for raising old questions about choice and competition. 

Mozilla also operates Gecko, one of only three major browser engines alongside Google’s Chromium and Apple’s WebKit. That gives Mozilla a key role in shaping how open web standards are developed and implemented through groups such as the World Wide Web Consortium.

It hasn’t all been smooth sailing. Mozilla’s also had setbacks over the past year or two. In late 2024, it announced plans to lay off around 30% of its staff and last year it shuttered products like Pocket as part of a plan to refocus on offerings.

Finding moonshots on Earth

Mozilla’s new report is more like a manifesto designed by an “underground collective” inspired by punk and resistance movements of the 1970s and 1980s. The microsite’s design seems to intentionally reject the minimalist uniformity common with Big Tech brands and rebrands.

Mozilla’s efforts also include a new “Choose Your Future” campaign for internet users, developers and advocates interested in charting a new path. The campaign is anchored by a series of five short videos that’ll be featured on social media and through ads on platforms like Reddit, Meta, and X.

The ads all have different messages, but the same ending sound: a modem dial-up as a nod to the internet from a few decades back. Each features a dystopian parable for an AI era without options but with plenty of AI slop and intrusive chatbots. One video starts with a girl staring at a toy called “Funblock,” which a radio ad markets as “the only block you’ll ever need.”

“No choices, no options, no confusion. Just endless identical fun,” the narration says. “Funblock may result in boredom, diminished agency, and loss of independent thought. Ask your algorithm if fun is right for you.”

Mozilla’s new AI strategy exists in an uneasy tension of how to build trustworthy tech in an industry obsessed with growth. Can it offer a viable alternative to Big Tech’s tightly integrated ecosystems while still being the internet’s moral compass? 

Surman thinks so, adding that Mozilla’s having the same AI debates internally as the world is having outside it: what to do with AI, what not to do, when it’s useful, when it’s scary, and how to make tech that’s better for everyone. But instead of putting data centers on the moon, Mozilla hopes to forge a future that’s privacy-enhanced, open-source, cheaper, and more environmentally friendly.

“[People say] ‘You’re crazy, that can’t happen,’” Surman says. “But you think we’re crazier to do a collective barn-raising for something that is joyous and great, and you’re going to put data centers on the moon, and we’re the ones who aren’t grounded in reality?”