After backlash over Firefox AI plans, a dev clarifies users can fully disable AI features
Following the recent backlash over Mozilla’s new CEO and his previously announced stance on AI features in Firefox, a Firefox developer named Jake Archibald issued a public clarification on Mastodon addressing some of the most common concerns raised by users. He explained that the browser will include a global AI kill switch that allows users to fully disable all AI related features across the browser. Once enabled, the switch removes AI elements from the interface and prevents them from returning unless the user manually turns them back on.
Archibald also confirmed that all upcoming AI features will be opt in, meaning users will not be enrolled by default. He acknowledged that opt in can be interpreted differently, such as whether a toolbar button appearing by default counts as participation. Even so, he emphasized that the kill switch is definitive and overrides any ambiguity by completely disabling AI functionality.
While the clarification helps ease privacy concerns and makes it clear that AI features will not be forced on users (at least for now), many continue to argue that the core issue is not whether these features can be, but that development resources should instead focus on performance and core browser improvements, and that the very decision to invest time and money into AI is seen as a clear signal of Mozilla’s priorities. In parallel, other browsers have used the situation to speak up and clearly state their opposition to AI integrations, including the also Firefox based Waterfox and Vivaldi.



Comments
Even if it is a pure AI kill switch not tied to core features (doubt), it is still unnecessary bloat. I'll stick with lean and privacy-focused forks.
Opt in is not for interpretation, it is very clearly defined. If you need a killswitch to "disable" it via a button, that means it was on by default, making the killswitch opt out, definitionally. If they really wanted to have the trust of people and have it actually be opt in, you would not implement it ever in the code of the browser and only download it after the user explicitly pushes a button to consent to it. They are backpedaling and panicking. I switched to Floorp, with Waterfox as a second alternative who have promised to not have any AI features.
I have no doubt that they'll pull a Google and tie core features to AI features such that if the “AI kill switch” is used, things people regularly use either won't work at all or won't work like they used to.
Why not just make them plugins?
Thank you! We needed to know this was optional and if it's ever not WaterFox get's install! Maybe if the AI were purely assistive I'd be okay, but you do realize that AI is incredibly new tech with a lot of flaws and should be pursued only when definite improvements can be made using such technology.
This should be the end of any new technology in general, if it's not going to cause immediate improvements for the User.
Even if they AI is trained and runs ethically, it must first be vetted by test users for a good while.
AI might have some good things going for, DotComs are still used, just pushing technology for no reason sucks!
Like I said once you slop, you can't stop. As soon as Mozilla shove slop inside Firefox, it became irrelevant as if the intention was to kill Firefox all along (1, 2, 3.)
Mozilla's soul, who once respected the Manifesto, has been sold to GAFAM so move to a de-enshittified Firefox forks or use Gemini (protocol) instead since the original freedoms of the free web is slowly being stripped away by proprietarists (GAFAM with Chrom* (1, 2, 3,) and Enshittified Malware Extensions by copyright extremists.)
Do not let proprietarists achieve their total victory by making ALL devices run GAFAM software. This is what this site is for: to find "alternatives" (sic) to GAFAM software. And once again, Stallman was right: software freedom means self-control, something which proprietary malware kept on suppressing it (control without any accountability,) which means it is a culture of resistance against GAFAM and any entities whom their evangelist infiltrated.
Sure, AI is opt-in until the day Mozilla will complain not enough people are using it.
But if the AI is not smart enough to re-enable itself alone, users must be safe, so far. But given the recent announcements, AI may be already replacing the current Mozilla CEO.