Firefox 148 lets you disable ALL AI features with one click

Let me break this down: imagine that every time you open your browser, someone has already decided which features you need. Automatic AI translations, built-in chatbots, tabs that organize themselves, summaries nobody asked for. That's what Chrome and Edge do right now. Firefox just said: enough.

On February 2, 2026, Ajit Varma, head of Firefox, published a post on the official Mozilla Blog announcing something that had been brewing for months: Firefox 148 will include an "AI Controls" panel with a master toggle that disables all artificial intelligence features in the browser. One switch. Everything off. No questions asked.

It's the first time a major browser offers anything like this. And it's no coincidence that it arrives at a moment when only 35% of people trust AI, according to a 2025 Edelman survey. AI fatigue is real, and Mozilla knows it.

In this article, I'll walk you through what this kill switch actually does, which AI features it blocks, how it compares to what Chrome, Edge, and other browsers offer, and why this decision might be Mozilla's smartest move in years.

What is Firefox's AI kill switch and how does it work

The master toggle

The trick is in its simplicity. In Firefox 148's Settings, you'll find a new section called "AI Controls". Inside, there's a main toggle: "Block AI enhancements". Flip it on and you're done.

When it's activated:

• All generative AI features are disabled
• You won't receive pop-ups or reminders about AI features
• Future AI features that Mozilla adds will also be automatically blocked
• Your preference persists across browser updates

But Mozilla didn't build this as a sledgehammer. If you want to keep some features and block others, you can. Below the master toggle, there are individual controls for each AI feature.

The 5 AI features you can block

Think of it like a house with five smart appliances. The kill switch cuts the main power, but you can also choose which ones to leave plugged in:

What most guides won't tell you is that Firefox's translations run entirely on your device, without sending data to the cloud. It's one of the most privacy-respecting AI features out there. Blocking it with the master kill switch is like throwing the baby out with the bathwater. That's why the individual controls matter.

When it arrives and how to try it now

Firefox 148 launches on February 24, 2026. But if you want to try it right now, you can do so in Firefox Nightly:

1. Download Firefox Nightly
2. Type about:config in the address bar
3. Search for browser.preferences.aiControls
4. Change the value from false to true
5. Go to Settings → you'll now see the "AI Controls" section

Default state: The toggle comes off (AI active). Features you've already used appear as "Enabled," features you never touched show as "Available," and features you manually disabled show as "Blocked."

Why Mozilla did this: the story behind the kill switch

The December disaster

To understand this move, you need to rewind two months. On December 16, 2025, Mozilla's new CEO, Anthony Enzor-DeMeo, published a blog post declaring Firefox would become a "modern AI browser" and grow "from a browser into a broader ecosystem of trusted software."

The internet erupted.

On Reddit, an open letter challenged Mozilla's narrative. Comments were merciless:

"I've never seen a company so astoundingly out of touch with the people who want to use its software."

Within 24 hours, Enzor-DeMeo had to respond personally on Reddit under his anthony-firefox account:

"Rest assured, Firefox will always remain a browser built around user control. That includes AI. You will have a clear way to turn AI features off. A real kill switch is coming in Q1 of 2026."

Jake Archibald breaks the silence

Jake Archibald, Web Developer Relations Lead at Mozilla, was more direct on Mastodon:

"Firefox will have an option to completely disable all AI features. We've been calling it the AI kill switch internally. I'm sure it'll ship with a less murderous name, but that's how seriously and absolutely we're taking this."

And he added:

"The kill switch will absolutely remove all that stuff, and never show it in future. That's unambiguous."

The Mozilla pattern

If you've been following Mozilla for years, this cycle should sound familiar: controversial announcement → massive backlash → backtrack with user controls. We saw it with the Mr. Robot extension in 2017 (silently installed on users' browsers), with the Pocket integration, and with terms of use that seemed to grant Mozilla a license over user data.

Mozilla isn't malicious. But it has a constant tension between its privacy mission and its need to find revenue sources beyond the Google deal (which accounts for over 80% of its income). AI is a clear bet on diversification, but its user base is precisely the one that most rejects forced AI.

Firefox vs Chrome vs Edge: the war over AI control

Chrome: forced AI with no escape

Google doesn't beat around the bush. In October 2025, it forced Gemini and AI Mode into Chrome with no opt-out. Chrome's AI features include:

• Gemini integrated into the search bar
• Automatic AI tab organization
• AI theme generation
• "Help me write" (writing assistant)

The controls are fragmented and incomplete. You can disable some things, but there's no master toggle. And Google keeps adding AI features with every update without asking.

Edge: Copilot everywhere

Microsoft is even more aggressive. Edge has:

• Copilot Mode: transforms the entire browser with AI
• Permanent Copilot sidebar
• Multi-tab reasoning
• Automatic content summarization

Microsoft claims Copilot Mode is "opt-in," but the reality is AI features are everywhere and hard to fully disable. A Change.org petition demands Microsoft stop forcing Copilot into Windows and Edge.

The irony is exquisite: CEO Satya Nadella published an essay asking people to stop calling AI "slop" (garbage). His own Copilot analyzed the text and concluded it "sounds like a robot wrote it." And according to internal leaks, Nadella admitted Copilot integrations with Gmail and Outlook "don't really work."

Brave: opt-in AI from the start

Brave took a different path. Its Leo assistant exists, but:

• It won't activate unless you invoke it
• It doesn't log data or train models on your queries
• Your IP is hidden via reverse proxy
• No account required

Brave proves you can offer AI without forcing it. The issue is its market share remains marginal.

Safari: system-level AI, not browser-level

Apple delegates AI to Apple Intelligence at the OS level. You can disable Apple Intelligence entirely from System Settings, and Safari simply respects that decision. It's a clean approach, but only works within the Apple ecosystem.

Comparison table

AI fatigue in numbers: why millions want an off switch

This kill switch doesn't exist in a vacuum. The data explains why so many people want to disable AI:

• Only 35% of people trust AI globally, down from 50% in 2019 (Edelman, 2025)
• 54% of people are wary of AI systems (KPMG, 2025)
• 42% of companies have scrapped the majority of their AI initiatives, up from 17% in 2024 (S&P Global)
• 80% of workers using generative AI say it added to their workload (Upwork)
• 45% of frequent AI users report burnout, versus 35% of non-users (Quantum Workplace)

What most guides won't tell you is that AI fatigue isn't just a consumer thing. Half of business leaders report declining enthusiasm for AI, according to an EY survey. The hype cycle is giving way to realism.

Firefox isn't offering a kill switch because it's anti-AI. It's offering one because it understands that user control is a product in itself. In a market where everyone forces AI, the option to turn it off is the differentiator.

What they're not telling you: criticisms of the kill switch

Opt-out is not opt-in

The sharpest criticism is that the toggle comes off by default. That means AI is active until you go to Settings and turn it off. For privacy purists, this is insufficient:

"This is clearly opt-out. If Mozilla can't agree to that basic definition, I don't see how users are supposed to trust it'll actually work."

The argument is valid: if you truly prioritize user control, the default should be AI disabled, and the user enables it if they want.

"AI enhancements"? The language trick

The toggle says "Block AI enhancements." Calling something you want to disable an "enhancement" is what UX designers call confirm shaming: making the user feel like they're missing out on something good.

A more neutral label would be "Block AI features." The nuance matters: one is disabling features, the other is "blocking enhancements." The latter implies you're rejecting something beneficial.

Translations: collateral damage

Firefox's translations are arguably the most useful and least invasive AI feature in the browser. They work 100% on your device, without sending data to any server. Blocking them with the master kill switch is disproportionate.

Mozilla seems aware of this, which is why they offer individual controls. But the master switch's message is "all or nothing," which doesn't reflect that some AI features are reasonable and some are questionable.

Marketing or conviction?

Some skeptics see the kill switch as marketing theater: Mozilla gets positive press for offering the toggle, while still shipping AI features enabled by default. It's the corporate version of "you can say no" while pushing you to say yes.

On the other hand, no other major browser offers even that. Criticizing Mozilla for not being radical enough when Chrome and Edge don't even give you the option is missing the forest for the trees.

Firefox and its market share: last desperate play?

There's an elephant in the room we can't ignore: Firefox has a 2.4% global market share. In 2009, it had 31.82%. The decline has been brutal.

Edge has already surpassed Firefox. In terms of users, Firefox has between 305 and 362 million, but the trend is downward.

In this context, the kill switch isn't just an ethical decision. It's a strategic one. Firefox can't compete with Chrome on features or Safari on ecosystem. Its only viable differentiator is privacy and user control. If it abandons that, it loses its reason to exist.

The kill switch is, in a sense, Firefox saying: "We're the browser for people who want to decide for themselves." It's a niche, yes. But in a market of 4.5 billion internet users, even a 2% niche is hundreds of millions of people.

What else comes in Firefox 148

The kill switch grabs the headlines, but Firefox 148 (February 24) includes other technically relevant additions:

• WebGPU in Service Workers: Enables accelerated graphics processing in the background
• Sanitizer API: Cleans potentially dangerous HTML before inserting it into the DOM
• Trusted Types API: Prevents cross-site scripting (XSS) attacks
• Screen reader improvements: Better compatibility with accessibility technologies

These are technical improvements that benefit developers and users alike, without forcing anything or asking you to trust an AI model.

FAQs: Frequently asked questions about Firefox's AI kill switch

When will the kill switch be available?

The kill switch ships with Firefox 148, launching February 24, 2026. If you can't wait, you can try it now in Firefox Nightly by enabling browser.preferences.aiControls in about:config. In the stable release, it will appear directly in Settings → AI Controls with no manual configuration needed.

Does the kill switch block absolutely all AI?

Yes. According to Jake Archibald at Mozilla, the kill switch "will absolutely remove all that stuff, and never show it in future." This includes current and future features. If Mozilla adds new AI features in Firefox 149, 150, or later, they'll also be blocked if you have the toggle enabled.

Does it affect browser performance?

It shouldn't. Firefox's AI features are modular, and disabling them doesn't affect browser performance. In fact, by blocking background AI processes (like smart tab grouping), you might notice slightly lower memory usage.

What's the difference between this and using LibreWolf or Waterfox?

LibreWolf and Waterfox are Firefox forks that strip out telemetry and controversial features from the ground up. The difference is that with Firefox 148, you can have a browser maintained by Mozilla (with frequent security updates and enterprise support) plus AI control. With the forks, you lose official support but gain a more radical approach from the start.

Will Chrome or Edge offer something similar?

No signs so far. Google and Microsoft see AI as central to their product and business strategy. Chrome integrates Gemini ever more deeply, and Edge bets everything on Copilot. Offering a kill switch would mean admitting AI might be undesirable, which contradicts their commercial narrative.

Conclusion: the browser that lets you choose

Firefox's kill switch isn't perfect. It comes disabled by default (opt-out, not opt-in), uses language that favors AI ("enhancements"), and arrives after a considerable communications stumble. Mozilla has a track record of promising privacy while looking for ways to monetize, and that tension doesn't disappear with a toggle.

But here's the reality: it's the only major browser offering you this option. Chrome forces Gemini on you. Edge stuffs Copilot into everything. Safari delegates to Apple Intelligence. Only Firefox says: "Here's a switch. You decide."

At a time when only 35% of people trust AI, when 42% of companies are abandoning AI projects, and when big tech CEOs themselves are starting to admit their integrations "don't really work," the right to say "no thanks" isn't a whim. It's a necessity.

Firefox 148 arrives February 24. The button will be there. The question is whether you'll press it.