Share your feedback on the AI services experiment in Nightly

[asafko]

Hi folks, 

In the next few days, we will start the Nightly experiment which provides easy access to AI services from the sidebar. This functionality is entirely optional, and it’s there to see if it’s a helpful addition to Firefox. It is not built into any core functionality and needs to be turned on by you to see it. 

If you want to try the experiment, activate it via Nightly Settings > Firefox Labs (please see full instructions here). 

We’d love to hear your feedback once you try out the feature, and we’re open to all your ideas and thoughts, whether it’s small tweaks to the current experience or big, creative suggestions that could boost your productivity and make accessing your favorite tools and services in Firefox even easier.

Thanks so much for helping us improve Firefox!

[idcjustletmein]

I checked my browser settings to turn it off, and saw that it was on by default, which is the opposite of what you claim here. Generative AI has well-documented issues with intellectual property rights, the environmental impacts, and literally doing people's thinking for them, which I doubt I need to explain any further than it has been already. Mozilla is making its product worse to chase a harmful trend, and abandoning its core values by doing so. I will immediately be looking for alternatives and/or taking the advice of the other commenters who explained how to revert to a previous version.

If Mozilla needs another reason to revert this change (and the changes to the Privacy Notice, while they're at it), I am currently addicted to AI chatbots. They have isolated me from people I care about, made it harder to brainstorm creative writing ideas (which is something I love to do!), and wastes a ton of water while I generate answer after answer, and for what? To waste my own time pretending I'm dating a fictional character when there's many wonderful people I could be getting to know in the real world?

AI addiction is not something medical professionals are currently equipped to treat, and Mozilla has ignored all these issues, dismissed the impacts on people's health and safety, for a quick buck. Is that it, Mozilla? I can't trust myself to manage this addiction yet, and I'm only just starting to acknowledge the problem. I can't trust myself to manage a currently untreated addiction with a simple checkbox. There was a news story about a teenager who killed himself because of Character.AI, and you want to make the problem worse?

Whoever made this decision is gonna get a reality check from Firefox's userbase VERY soon. We chose Firefox because of its commitment to respect the privacy of its users, not sell our data or let companies track us... and now our data is being used to train a technology that already has blood on its cold, mechanical hands.

Well, Mozilla? You wanted people's data, didn't you? Or was this too personal, even for you?

[Manchester100]

A thoughtful and helpful response, I believe. Thank you.

[Kissaki]

> When you choose a chatbot you agree to the providers terms of use and privacy policy.

How am I supposed to make that assessment and decision?

At least links to those products would help. Some guidance or description would help too.

Supporting local AI (ollama) would also be helpful.

[Ginny]

I feel personally insulted that Mozilla is considering AI integration in Firefox.  This,  along with the very worrying changes in their privacy notice that seem to conflict with their recent blog posts, makes me fear for the future of the browser.  If AI does take any step forward,  I will switch to another browser entirely.  Disappointing from a company I once trusted. 

[majesticmini471]

I completely agree! The internet is a stab in the back, now known as artificial intelligence.

[illusoryrealm]

hi! I made this account specifically to say:

its crap get rid of it

[Manchester100]

Well lots of people are leaving feedback... 3,721 replies so far.

Anyone know if they will respond or how many responses they need before they do?

[lackey]

I don't think they were looking for any feedback that wasn't positive. The rest of us are just getting softened up for the inevitable

[Clawz]

If this is further implemented into Firefox I will stop using this browser

[majesticmini471]

I so sadly agree, and I am a strong team member to Mozilla. Lord, so help us!

[IlIllIlllllIllI]

I literally made an account on this forum specifically to voice my discontent about this. Lastpass isn't built into my browser, neither is bitwarden. Not even Ublock origin. But I have to have a sidebar for the 2025 edition of smarterchild from MSN messenger?

LLM chatbots aren't **bleep**ing novel; they don't do anything valuable in the industry use cases they're put in, they can't do even the basics when you contact an actual business, and to top it all off they blow infinitely more carbon into the atmosphere than a human, for the sake of replacing a human.

Just NO. I don't WANT this in my browser. I don't want to have to turn it off and constantly revoke consent from hallucination engine #200 that only stops hallucinating after enough people have been burned by it.

Take the revocation of consent as a NO, don't press yourself in deeper. Holy. **bleep**ing. **bleep**. Guys.

Take. Revocation. Of consent. As a NO.

[IlIllIlllllIllI]

No.

[sp09]

Please please please stop making AI searches default. AI uses horrifying amounts of fresh water and energy and is trained on stolen ideas, articles, books and art. And, more often than not, it is not even correct or helpful. Just let us have a regular search engine.

[ikkqq]

sir, Where is Microsoft Copilot and Deepseek AI?

[lackey]

Maybe money has to change hands before Mozilla adds it

[offline_zh]

Please add DeepSeek or locally deployed API config for chatbot. Now it can be added via modify provider from about:config, but the context menu is not working.

[beckyricee]

The AI services experiment in Nightly has been insightful and user-friendly, offering quick responses and intuitive interaction. It handled queries like kuwait civil id enquiry efficiently, showcasing its practical utility. Overall, it’s a promising tool with great potential for everyday tasks.

[stacyjasper]

Hello,

I've been exploring the new AI services experiment in Nightly, and I think it's a promising step toward integrating useful tools more seamlessly into the browser experience. The sidebar approach feels non-intrusive, and the opt-in nature is reassuring for those concerned about control and privacy. While it's still early, I can already see potential in making everyday tasks quicker, especially for users who juggle multiple services or research frequently. I encourage everyone to give it a try through the Firefox Labs settings and share your impressions—your feedback really helps shape what comes next. Also, if you're diving into AI tools, I highly recommend checking out Anya Longwell for inspiration on how productivity and creativity can truly come together.

Thanks!

[kjoatmon]

I'd like to see a sidebar disable option in settings so I don't have tto go in, remove all this crap I will not use, then remove the menu item from the toolbar. Then, if anything else like this gets added, it will automatically be disabled because the sidebar is disabled. Also, a disable all AI or SI blocker would be nice. One that sutomatically adds -ai to all searches, filters out AI results, etc.

[Agentvirtuel]

Hello

About, sidebar.revamp.
https://connect.mozilla.org/t5/discussions/sidebar-and-vertical-tabs-launch-in-release-136/m-p/94547...

About, remove the menu item from the toolbar.

If you wish to test userChrome.css

#viewSidebarMenuMenu{display:none!important;}

https://www.userchrome.org/how-create-userchrome-css.html 

https://connect.mozilla.org/t5/ideas/ability-to-edit-context-menus-remove-unnecessary-options/idi-p/... 

About, others items.  **
https://support.mozilla.org/en-US/questions/1518067

** Your browser is being managed by your organization, that's normal.

[fool]

Looks like you went ahead and added it anyway. I think this was a bad move, considering how many people here are saying they don't want AI and they find it hard to trust companies that push AI.

[Proxiehunter]

My feedback is that AI does not belong in Firefox and the biggest improvement would be to remove any LLM based models that have been implemented in Firefox and never touch AI again. I remember you guys being smart enough to stay out of the NTF bull and the crypto currency bull so why are you suddenly falling for this AI bull?

[Shriradhakrishn]

I appreciate your recent innovation in integrating an AI sidebar on Firefox desktop. I would like to request that this feature be extended to Firefox for Android as well.

Given the increasing reliance on AI tools, having an AI assistant sidebar on mobile would greatly enhance productivity and user experience. Bringing this feature to Android would reinforce Firefox’s leadership in user-focused, privacy-respecting innovation.

Thank you for considering this request.

[mc4net]

OOoooo, what a shame, the privacy model is slowly disappearing into the general revenue model and thus losing all confidence that there is a better ethical idea behind it 🤔

[Devilpants]

lmao you completely ignored the literal thousands of posts worth of negative feedback about this and implemented this absolutely unwanted absolute trash anyway. If these are the people in charge of firefox, I'm switching browsers. Done with this short-sighted trend-chasing techbro nonsense. I hope when the genAI bubble pops, you're all left with nothing and have to get real jobs.

[SparkyUpstart]

I have no desire to try AI services, because AI services nothing other than environmental damage and content theft. It wastes resources, wastes time, and wastes my opinion of Mozilla. I've been preaching the quality of Firefox, Thunderbird, and the VPN for the past calendar year because I believed in it. Now? Now any time a friend or colleague asks me which browsers don't have meaningless bloatware and sinister tracking built into them, I'll be forced to tell them to not bother changing because Mozilla is throwing away their reputation to ride the coattails of Google's bad decisions, which is baffling considering so many people (myself included!) switched over BECAUSE of Google's bad decisions.

[Deciare]

With the closure of HuggingChat in July 2025, I've lost access to models I previously used. Enabling and configuring the localhost option has brought back the capacity to use some of those models, but there's only so much that my laptop can handle.

Are there plans for adding more chatbot provider options to the sidebar, or enabling users to multiple custom URLs as options?

[Deciare]

One of the weaknesses of this implementation, compared to Orbit, is that page context must be manually selected before the LLM can act on it. Although I could manually select all text on a page prior to querying the LLM about content on a page, that's an extra step and potentially less comprehensive than the way Orbit was automatically sending page context to the model.

I think it may be sensible behaviour to automatically extract page context if no text is selected at the time that select any of the chatbot functionalities from the context menu.