Isn't it odd? All those tech CEOs tell us that we won't be able to live in a world without AI, how AI will be within every single app, service or codebase eventually...

And then they constantly try to shove it into their products, with no way to disable it. I'm assuming the user data would show that quite a lot of people would turn it off, so to not ruin your own statistics for the next shareholder/investor meeting, you need to force them

> I'm assuming the user data would show that quite a lot of people would turn it off

You would be wrong — outside of the Hacker News bubble very few people mess with their default settings, in any app.

How can you justify such a wildly conclusive statement without providing any supporting information? It's not just the "HN Bubble", recent articles on The Register discuss how there is no way to guarantee the information provided to the 3rd party LLM servers in the prompting is free from further disclosure. This raises a host of concerns for professionals with a non-deligable duty to safeguard client/patient information from disclosure. To the point that if LLM captures that information, by definition the attorney or physician has disclosed client/patient information putting their license to practice law or medicine at risk.

Exacerbating the matter is there is zero disclosure on Mozilla's part detailing exactly what information is sent to 3rd party servers as part of its AI rollout in FF 145. Would you risk your license on some FOMO AI rollout in FF that, unless you, as the lawyer or doctor, have stepped though each line of the tens of thousands of lines of FF code associated with this new AI to meet your ethical obligation and answer the bar or medical board's inquiry on whether client/patient information has been sent to a 3rd party?

Without the ability to completely disable and turn all AI submissions off, Mozilla's "Trust Us" position doesn't allow anyone with such a duty to meet it. This is before you even get to confidential and proprietary or trade-secret type information applicable in any professional setting.

These are all vexing questions from a legal standpoint.

Even if that's the case, I doubt usage rates are very high.

On LibreWolf 145.0.1-2 those options are still available. I set all to false. But for these AI options and many others and also not only for Firefox/LibreWolf, you do need to block the connections. Don't rely on disabling options because most will not respect that. If Firefox goes this path and break LibreWolf, we need to switch to a decent browser and there are a few, still blocking connections for Google and Mozilla, etc. If you're not able to block connections with a firewall, at least install a Pi-Hole so you can block ads and custom domains like Google and Mozilla and Fakebook crap.

Firefox 145 here, just updated to latest this morning.

My about:config settings still disable the stuff. I get no AI Context Menu.

That's strange, which OS? I am on Arch and also on 145 and I get the "Ask an AI Chatbot" in the context menu. The settings used to work in the past so I am not sure what's going on.

I believe these are all the settings I have disabled for AI:

browser.ml.chat.enabled

browser.ml.chat.menu

browser.ml.chat.page

browser.ml.chat.page.footerBadge

browser.ml.chat.page.menuBadge

browser.ml.chat.shortcuts

browser.ml.chat.sidebar

browser.ml.enable

browser.ml.linkPreview.enabled

browser.ml.pageAssist.enabled

browser.tabs.groups.smart.enabled

browser.tabs.groups.smart.userEnable

browser.tabs.groups.smart.userEnabled

extensions.ml.enabled

sidebar.notification.badge.aichat

Am I missing anything?

Seems whatever I had disabled earlier is still disabled on my install of FF 145.

I do have these additional settings.

browser.ml.chat.maxLength=0 browser.ml.chat.prompt.prefix="{}" browser.ml.chat.prompts.0="{}" browser.ml.chat.prompts.1="{}" browser.ml.chat.prompts.3="{}" browser.ml.chat.prompts.4="{}" browser.ml.chat.shortcuts.custom=false browser.ml.linkPreview.longPress=false browser.ml.modelHubRootUrl="example.com"

As far as I can see, that's it. Or at least I'm not seeing anything else related that I've disabled.

I had to go out. When I'm back home in a few hours, I'll try to look up all I've disabled.

Surely they're blowing substantial cash on this, right? I'm not sure what sort of cost/benefit analysis is convincing every last tech company to fit the bill for a gimmicky AI add on.

You still have the option to disable the AI assistant. You simply need to disable the relevant item in about:config.

Ah...you ask which ones need to be disabled? Find it yourself. You have the right to disable it.

> The Mozilla Manifesto: Principle 9

> Commercial involvement in the development of the internet brings many benefits; a balance between commercial profit and public benefit is critical.

Isn't that context menu the one that is disabled from within the context menu itself? Or have they removed that option?

I just did this now. I can't remember exactly what it said, but something like "Remove AI chatbot" and I clicked it, and it was no longer in the context menu. Just after updating versioh 145.0.1 to 145.0.2.

If it can't truly be disabled is there a question one could ask AI that is not illegal but would get their IP perma-banned? I have a static IP is why I am asking. Or perhaps domains that could be blocked in Unbound. What irritates AI owners the most?

It's like cancer spreading everywhere. If you go on Ecosia, there's a button for "AI Search" but no button for normal search.

Thanks for reminding me to disable this in about:settings. Mozilla is doing everything it can to push me away from Firefox.

Has anyone verified if this is just a UI leftover, or if the underlying services are actually still active in the background even when disabled? I've been considering switching to a fork like LibreWolf if Mozilla keeps hard-baking these features into the core.

The “underlying services” are a web view in a sidebar that defaults to chatgpt.com or claude.ai.

If you don’t open that, I’d be surprised if anything was instantiated or sent at all.

Does any software have a survey for what users want? Instead of immediately pushing AI, they should have pushed a survey where AI was one of the choices/questions.

if they do that there's a risk of getting an answer they don't want

Showing a pop-up to a survey honestly seems more annoying than an unobtrusive “open LLM sidebar” context menu switch.

Firefox also collects usage data if individual features (which can also be disabled), so they’ll learn about the adoption soon enough.

It is either a fork or new software based on what a people really want or need. This is not an easy hill to climb.

Unfortunately, "configuration" is the survey. I detest both configuration and surveys. Modify the open source code and rip out the Artificial Inference code because Firefox is open source, or build software from scratch: servo, ladybird, your own web browser based on a survey.

they also removed the user facing setting to disable ocsp stapling - ocsp stapling leaks domain info. it can only be found in about:config now.

I’m confused about for this assertion, for two reasons:

- My understanding is that OCSP stapling stops leaks, because the browser can get OCSP data from the server instead of needing to fetch it separately.

- Last I heard, Firefox was in the process of removing OCSP responder checks (precisely for privacy reasons) in favor of CRLite-based revocation checks—are you sure they didn’t remove whatever setting you’re referring to from the UI because it’s no longer relevant?

Shh, you’re interrupting the weekly Mozilla/Firefox shunning ceremony of HN with your uncomfortable looking at the facts.