222 comments

[ 2.6 ms ] story [ 245 ms ] thread
About 17 years too late to be AI-first.
MMW: This will be the death of Chrome.

I also do not see Google parting with something so critical to their advertising. With their own browser they control the full length of the wire between the ad-server and the user. Without it, they don't. Only way I could see this happening is if Google then released what they considered a better browser.

> I also do not see Google parting with something so critical to their advertising.

This is not really their choice at this point. They were already found to have abused their position so it's up to a judge to decide what Google has to to next. Google doesn't need "a browser", they need a tool that allows them to exercise more control and this whole court case is about preventing that.

OpenAI is just looking for new ways to funnel data into the training of their models. And I'm afraid so many people would eat it up as long as OpenAI gives them some AI candy in return.

> I also do not see Google parting with something so critical to their advertising. With their own browser they control the full length of the wire between the ad-server and the user. Without it, they don't.

They've already been convicted of anti-trust behavior for precisely this reason. Now the trial is in the remedy phase where the DOJ is asking that they be forced to divest ownership of Chrome and other properties.

Google will have no choice in the matter. It's entirely up to the judge at this point.

Up to 'the' judge', plus the many other appellate judges, unless DoJ and Google come to an agreement and Google decides not to appeal. Google can both appeal the original verdict and any remedies.
Why not fork it like Microsoft did with Edge?
Cause the value is in the users. The actual browser is close to worthless.
Chrome/Chromium's value is in the development momentum. You can pile up features, manipulate the entire web, and make it impossible to compete for others. This is what they're trying to buy. A fork isn't enough - there's a huge difference between a technical fork and a meaningful fork.
That's a strength of Chrome for Google. OpenAI would not enjoy the same benefits as they don't control the whole vertical. They want the 4 billion strong spyware botnet that is Chrome.
You have to control a meaningful part of the web (like the Google ecosystem) for that to work.
(comment deleted)
Microsoft Edge is relying on Chromium being contiously developed by Google, isn't it? Not a hard fork that no longer receives updates from upstream.
I don't even use chrome and this sounds miserable.

I had to run it for something the other day and immediately got nagged to remove uBlock Origin because they automatically disabled it. And I'm just thinking.. I will never, ever use this browser for anything other than light dev work if I really needed to.

Using ublock origin lite, haven't noticed a difference. It still blocks all the ads including google's and youtube.
I think it could be interesting to see Chrome sold to ten or more different entities. Allow none of them to use the Chrome name. The last thing Google can do with Chrome is have it randomly select one of the new variations for updates.
Let's give Chrome to a million different entities, for free. None of them can use the name. We can remove all the Google add-ons as well. Call it something like... Chromium?
How would this even work in practice? Hundreds of million installs out there already use Chrome. How do they get updates (from servers) and which browser does it upgrade to?
The DOJ / FTC / W3C / SEC / whomever vets all of the entities purchasing a share of the Chrome userbase. Part of the purchase agreement would be to have the update and development infrastructure in place and commit to updating their version of the browser and making good faith efforts to adhere to W3C standards for the next 3 years.

Once the sale details are finalized, Google pushes out a final update that changes where the next update to Chrome would come from (and it would be a random selection from the list of buyers).

Make default search be a version of ChatGPT and put ads on it? Could work (I wouldn't use it though). The way a lot of people use their browser they might be fine with it if it puts navigation-type links up top (I have literally seen a technical colleague with a PhD do a Google search for "x" and click a link rather than type the ".com"). The inference costs would surely make it hard to profit from though.

Having just closed a $40 billion USD funding round [1], OpenAI might actually be able to afford a fair price for Chrome (supposedly $15-20 billion USD [2]).

[1] There are some catches to that: https://www.investopedia.com/openai-closes-up-to-usd40b-fund...

[2] https://duckduckgo.com/?q=chrome+sale+valuation

Ads? Customers would not pay for ads, they would pay for getting convenient "truths" emphasized in the training material, and inconvenient ones deemphasized. Imagine Wikipedia with pay-to-edit.
That's the beauty of ads. Customers don't pay, the advertisers do.
Customers are whoever is paying. Non-paying users are not customers, they are the resource to be sold.
If I'm typing website name, I'm always using Google to ensure that I won't mistype it and won't land on malware domain. Google always corrects me and will make sure that search results are safe.
Autocompletion is much better for that IME. SEO is too good for anything not really well known, so I find I have to pay much more attention with search results.
You can ignore it, this is just a page taken from the Musk school of attention farming.
It actually kind of makes sense.

The DOJ wants to break what it considers to be Google's monopoly, and Chrome is a prime target. The problem is that Chrome by itself is worse than worthless, it is a money sink and it only makes sense as a part of a system.

OpenAI is starting to feel the competition. ChatGPT is no longer the only game in town, DeepSeek happened, Google is becoming actually good, Claude is quite popular among coders, and Grok is not a joke anymore. They need something if they don't want to lose out, and buying the most popular browser to make it into a gateway into their service may be an option.

> Google is becoming good

Where?

New Gemini models are quite good, Gemini 2.5 Pro is 1st in the user-benchmarks [1]. They also have Gemma, very good model that can run locally [2]. Benchmarks are not oracles of truth, but I feel like Google is not a kid who arrived late at the party anymore.

[1] https://lmarena.ai/

[2] https://blog.google/technology/developers/gemma-3/

Yeah, with Gemini 2.5 Google stepped up to the grown up AI table and added on top. I still have a soft spot for Claude for general purpose chats, but have fully switched to Gemini for dev.
Considering that several key breakthroughs happened at Google that made GPT-style LLMs possible (e.g. Attention is all you need paper), it's more like they took a long smoke break than showed up late.
> The problem is that Chrome by itself is worse than worthless, it is a money sink and it only makes sense as a part of a system.

Or, just an out-there idea, what if the Chrome became property of the government instead? Forced to be FOSS, put into maintenance mode and offer it as a truly user-focused browser instead of driven by any for-profit company (which will eventually run it into the ground).

Considering the questionable choices made by the current President of the US let's maybe not do this. Let's not act like this wouldn't be turned into a propaganda machine in minutes
Yeah, let’s put Trump in charge of the most popular browser in the world. What could go wrong?
Yep, move it into an independent non-profit foundation and have the government fund its development through taxes as a public good software that benefits everyone. The idea makes sense, which is why it'll never happen.
> Yep, move it into an independent non-profit foundation and have the government fund its development through taxes as a public good software that benefits everyone. The idea makes sense, which is why it'll never happen.

Or turn it into a tightly regulated natural monopoly, a la a public utility.

But I totally agree with you: some things should just be state-owned. We should put our energies into identifying those things and addressing any legitimate concerns (e.g. spying via requiring open source and reproducible builds) instead of trying to free market all the things.

Especially during the current (2025) iteration of the US government. All sides of the political spectrum will be accusing the program of having political bias against each other.
> All sides of the political spectrum will be accusing the program of having political bias against each other.

That's pretty much the optimal state of the current iteration of democracy, isn't it? I'd feel more scared if everyone agreed to it and there was no complaints, then something is guaranteed to be fucky-wucky for the normal person, but without knowing about it. At least if both sides complain about the other, it's relatively impartial.

Chrome is already FOSS, and there is no shortage of forks, including Edge.

The only part that isn't is the brand, and the ties with Google. And I am not a fan of the idea of a (foreign in my case) government browser, I'd rather have Google. At least, Google has a presence in my country and is bound by its laws,

Chrome is not FOSS. Chromium is FOSS. Chrome is a proprietary fork.
> put into maintenance mode

That seems like a death sentence. The standards aren't stagnant.

(comment deleted)
But removing Chrome from Google makes zero sense and won't stop it from being a monopoly. The monopoly part comes from buying DoubleClick (in 2007!) -- that should never have been allowed.

Not sure how to extract that part from Google now. It would be difficult, but probably quite effective.

Quite effective at destroying anything good left of Google I would say.

Google has a bunch of nice things (search, gmail, maps, ...) that cost money, and an advertising business that makes money, the former helping the latter. Split the two and the nice stuff will be without funding and die out, and only the "evil" part will survive. Or so I think. Splitting out Chrome will not change the face of the world, but Firefox has shown that an (somewhat) independent browser can work.

Those 'nice' parts of the google are feeding the 'evil' advertisement business. Now more than anything, the reason google's ad business is so rich specifically because they (and only they) know everything about most of the denizens to farm them efficiently and thus demand a premium. Take their feed away and the ad business livestock will suddenly be lot more docile.
It would be interesting to hear how much a subscription to docs, gmail, maps, etc. would have to be to keep them operating at current levels.
Google could have had solid competition if Garmin and TomTom made their offline map subscriptions available on phones instead of pushing their dedicated devices so hard. TomTom just started publishing their own apps during Covid, and Garmin is still nowhere to be found. There's also other apps like Magic Earth that use other data, but they're also super recent.

I feel like the whole tech industry, especially the American part, really dropped the ball on this.

> But removing Chrome from Google makes zero sense and won't stop it from being a monopoly. The monopoly part comes from buying DoubleClick

Not only. Google controls a lot of user attention. See how many services they link together to serve you ads .... erm .... recommendations to make browsing better or something: https://x.com/dmitriid/status/1908951546869498085 And one of those services is Chrome

They could just split DV360 from Google Ad Manager and Google Ad Exchange.
Totally agree, I stopped visiting Ars Technica because a lot of the "journalism" is reports on Elon Musk and reposts from Hacker News. It is very clear some of their writers just watch for what is popular on this site then write about it (which is not bad in itself, they just don't put more effort than the original report on it).
The original Ars I had bookmarked and visited every day. With seriously in depth articles about computers. When they got bought out it quickly became attention seeking with very shallow articles. It has not gotten better. I had honestly forgot they existed.
> it quickly became attention seeking with very shallow articles

Welcome to the internet post-2008

This sounds a bit clickbaity, they could just fork Chromium and build their own version as Edge or Brave. After all, they already have the distribution (ChatGPT).
Surely they'd be wanting to buy 3.5billion users, though?

I actually wonder what the price tag is for that, lol.

That's it, if Chrome goes independent and / or is opened up for purchase, it's a very attractive target for companies for data harvesting. I mean I'm sure Google does as well (and got in trouble for e.g. incognito mode not being as incognito as advertised), but they keep it under wraps and make sure to not scare people off.
If they get chrome they can push updates to the already existing user base.

They’d be buying the user base.

but my main reason for using chrome is my google account
Is there any talk about how google account integration would change if they were forced to sell?
I guess most people use chrome because it is the default. Some people think I don't have internet because I don't have a chrome icon.
The default where? In Windows the default is Edge, in Mac is Safari and in Linux distros usually is Firefox. And somehow people prefer to download Chrome.
> The default where?

Android and ChromeOS

And it's very, very different from buying the chrome itself and getting >50% of market share overnight.
The title is clickbait. Most of openai's comments were about their desire to have access to google's index. They also discussed how open AI is thinking about creating a chrome Fork to compete with Google Chrome. The specific part where they mentioned wanting to buy chrome was a hypothetical muse:

> According to Turley, OpenAI would throw its proverbial hat in the ring if Google had to sell. When asked if OpenAI would want Chrome, he was unequivocal. "Yes, we would, as would many other parties," Turley said.

I think you’ve proved your own point: yeah they could launch the next brave, but then they’d have brave and not chrome. 99.9999% of the world haven’t heard of brave.
They do not have the distribution, that would be the entire point of buying Chrome. ChatGPT is not a web browser, buying Chrome lets them hoover up web browsing data + billion users, which are crucial to developing an ads product, which is ultimately what this will shake out into.
The Chrome brand (logos, trademarks, its "word on the street" recognition - its marketing) are the true meat and potatoes of that web browser. Forking Chromium would force them to build a new brand on that scale from scratch. That's not an easy thing to do.
Google is likely to have 1000 people or so working on Chrome and Chromium.

That's going to be difficult to maintain. If OpenAI takes over i expect Chrome and Chromium to go closed source.

Chromium would be hard to relicense due to copyright but trivial to fork.

So, chromium won't go away. Those 1000+ people are the main resource here. Effectively they work mostly on chromium and not on chrome. What happens to chromium if that stops?

My guess is MS might step up and hire people.

> "So, chromium won't go away."

Sadly, it absolutely can go closed-source as it's licensed under BSD-3 clause, which is not a copyleft license.

I think the old versions would still be OSS (or "source available" at the very least) and could be used in some way in other products.
The old version would stay licensed exactly as it is. You can't retroactively change a license.

Technically any form of re-licensing the original implies some kind of fork happens. The old system continues to exist for anyone that has a copy. The license allows anyone to fork of course. So anyone with a copy could just pretend that's the main fork.

Chromium is actually a complicated project. Some of the project coordination is actually taken care of by the Linux Foundation; not by Google. But Google of course controls and hosts the key code repositories. However, they don't own the code base exclusively though because they have been accepting contributions from e.g. Microsoft and the many other companies that base their browser on Chromium. And of course individual developers. Copyright stays with those.

The partial permissive licensing (it's a mixed license code base) makes it possible to incorporate the code in a closed source system. But some of the components are LGPL, which still would require publishing the code of those components. And of course Chrome is an example of such a system and Chromium is the way they publish the source code. And Edge. And Brave. Etc. Chromium is the shared source code base for those. But it doesn't allow Google (or anyone) to take away the copyright from contributors. That would require a copyright transfer agreement. And no such thing exists for Chromium.

So, Chromium is unlikely to stop existing. And if it somehow does, it would immediately trigger a fork with a different name created by all the stakeholders that would require such a thing.

The only question is who ends up employing the Google employees that currently provide most of the code contributions and who owns the Chromium trademark (Google currently). IMHO that's an unhealthy situation that should be resolved in any case. And with multiple wealthy companies using chromium, funding that should be no issue.

> So, Chromium is unlikely to stop existing. And if it somehow does, it would immediately trigger a fork with a different name created by all the stakeholders that would require such a thing.

> And with multiple wealthy companies using chromium, funding that should be no issue.

You're assuming tech companies aren't being cheap lately when, in fact, they are being cheap. We should assume that the mooching (using Chromium without contributing code or money) would go to 11.

I assume a condition of the sale would be a contract to pay Google $XX M/yr for support.
I assume a condition of the sale would be a contract to pay OpenAI $XX B/yr for the default search.
Not sure why the DOJ would permit that.
I mean a condition for Google to continue to exists and for the spinoff of Chrome to be economically viable. I don't care about both these outcomes tho.
800 on tracking, 150 on marketing and 50 on features that are useful for the user...
0.5 people (part timer) on bug fixes, they slow roll the fixes for any of the bugs that are tracking loopholes.
I wonder if OpenAI would monetize the browser primarily by collecting user data and advertising, or by pushing ChatGPT subscriptions.
I guess they could effectively copy Kagi's model at mammoth scale - offering a premium internet browsing experience with a 'personalised assistant'.

Easy to convince at least 10% of the users to sign in to their browser with a verified credit card to 'protect the children', and governments around the world would give you full support.

At that point, would be trivial for them to track browsing habits, and then to start offering personalised assistants which save you time and eventually cost money.

Pretty sure you could save money throuh having a huge botnet of computers to tap into, and a huge amount of data to help cache and standardise common requests.

didnt they consume the entire web to make their models? Buying chrome is an anachronism, like google buying godaddy.

ah nevermind, it's just billionaire pissing contest

They've consumed yesterday's web.

Why not get the user to pay the energy and processing bill for subsequent rounds?

Being able to track the habits of 3.5 billion users at source is probably quite useful, too.

Imagine using the largest browser user base as an unlockable botnet for scraping, built right into their browser itself.
I think you're missing what makes consumer companies valuable. It's all about distribution. They get way more data (usage, browsing, etc.), they get 3.5 billion users (this is the main thing) and they get to be the interface for all those people onto the web. I just don't think they can afford it.
Is anyone working on an Ai first phone?

I think at the least I should be able to have Ai interact with anything on my screen. And beyond that it could even code interfaces on the fly depending on the task.

Chrome is installed on (almost?) every Android phone. So they'd be buying much more than this.

Not a new "AI phone", which has to gain traction, find users, convince people to switch, compete in highly competitive (hardware( and duopolized (OS, Software) landscape.

I won't be suprised if amongst Android users, Chrome is one of the most installed apps - if only because many phones have it locked (i.e. its really hard or impossible to remove).

Maybe "Google Assistant" is installed more than chrome, IDK. But Chrome has the additional benefit that it is also installed on many iPhones. Sou Chrome would be a gateway into "making your iPhone an AI phone" too.

It almost sounds as if you believe this would be a good thing.
It kind of shows the state of Indie development over the last 2 decades that it's only the big players that can move mountains. Linux on the phone never happened, and it seems like we are all resigned to the fact that the future AI phone OS/AI browser will be made by the titans and not anyone else. Even if it came out of the open-source scene the titans just buy the damn thing.
I have the latest S25 (regular not the oversized +) and it does a lot of AI first like things. It can see what's on the screen, summarize your day, circle to search etc.

I disabled most of it within a few days because it mostly gets in the way of normal basic things like taking screenshots or just reading my actual notifications in full.

The picture editing can be nice, but realistically there's just no need for most of its 'support', it's just clippy on your phone getting in the way.

I'd like this if it wasn't an avenue for all of my personal activities to be analyzed and sold.
Better get off the Silicon Valley hype train then, because every single company in it is valued on the basis of how complete their user surveillance panopticon is.
I don't know what's going on here but it sounds like "Hey, I want to take your wife and have sex with her". Gotta read between the lines here. I think Google has a 14% stake in Anthropic. Along with Gemini, Chrome and Android as delivery vehicles, and search. The monopoly lawsuit is about this (the advertising ship sailed long ago, so what's this really about), and there's some nasty legal talk going on here. I think if they just give up the Anthropic stake and promise to allow any AI provider in chrome, then this nonsense will all end.

/tinfoil

I am a big fan of having deep-pocketed corporations pay huge amounts of money to pay for open source products. I suppose this fits into that category.
> deep-pocketed corporations pay huge amounts of money to pay for open source products.

There are many names for this: co-opt, assimilate, bribing ...

A lot of times it is like when Tony Soprano offers you a deal, or like when the U.S. made the NAFTA deal with Canada and Mexico.

It feels good and awesome at the beginning but later on, when you become dependent on it, you'll have to pay an heavy price.

This is a transaction between 2 giant corporations. What is there to be a fan of? Chrome is not open-source, Chromium is.
Rony Abovitz, the founding energy behind Magic Leap, does this 'AI/XR Podcast' and last episode they discussed OpenAI rumors about a Social Network. As a founder himself who lived through being flavor of the day getting showered with venture money his observation was prescient: this 'I can do anything' approach is what happens to a certain kind of person (who will "eat everything on the table" is how he phrased it) given all checks and no balances.

He contrasted that with someone like Jenson who pulled Nvidia off of a cliff more than once and so has the scartissue to limit his reach to keep focus on core business.

> given all checks and no balances.

Do you mean no checks and balances? A check is a restriction or constraint.

I suspect this may be a play on words where checks = cheque, or money
"Balance" also has a financial meaning…
I am stating the obvious to clarify: cheques without balances is an overdraft. You give a piece of paper that can't be exchanged for money when promised.
I don’t think that the saying has anything to do with money. It’s about power, oversight and preventing overreach. Of course, that doesn’t mean it wasn’t used in reference to money by the OP.
Yeah I thought this was possible too that's one of the reasons I asked

I checked to see if that pun occurs elsewhere and didn't see it. Someone who doesn't have English as a first language may not know the more obscure usage of check since you don't use it much these days other than as part of phrases and idioms like "checks and balances", which is 18th century English

> scar tissue*

I got excited first thinking I’m starting my day off right by learning a new word, but nope :-)

> eat everything on the table

OpenAI probably senses they're not making ASI anytime soon. They have enough money to will themselves into a FAANG by essentially minting consumer and enterprise products. That could secure their long term future and returns.

Prediction: they burn through Microsoft’s loans buying social networks and browsers, and then when Microsoft stops writing checks microsoft acquire what’s left
Or someone else. They could definitely go the acqui-hire route.
I suspect that was the plan all along. I do not buy that these people believed in their own spiel for one second.
> the founding energy behind Magic Leap

So a scammer?

With which money, Sam?
Just shave off a few billion from that 7 trillion dollar plan. What's the big deal?
I see a great future for Firefox.
Chrome: unilaterally disables uBlock Origin

Mozilla response: mess around with Firefox's privacy notice in such a way that it generates _negative_ press

Potential future Chrome: gets bought by OpenAI

Estimated future Mozilla response: "every time a user installs Firefox, a healthy tree is chopped down, the wood is used to create bats with the user's name engraved on them, and the bats are used to hit endangered animals"

I love Firefox but Mozilla seems determined to self-sabotage itself. it is painful to watch
why do the not just make their own GPT browser?
Honestly, Google should sell Search to OpenAI. And keep Chrome and the rest of Google. It will be contrary to what the DOJ intended. But it makes sense for Google and Open AI.

By making search an AI first experience, both behemoths will signal the new dawn of AI is here.

Google’s greatest advantage is the use of AI in drive and docs and presentation and excel and cloud services.

AI search results (not just Google's) are so frequently incorrect that this would be disastrous.
Impaired assets are still assets, although it would be funny to see what fair market value would be for a busted search engine.
> Google should sell Search to OpenAI

OpenAI can't afford to buy Search.

Buy stadium naming rights and we can call the high point for OpenAi.
ugh. Would much prefer they do something separate.

The Chrome & derivatives mono-culture is going to become a problem down the line.

So if Google sold off "Chrome" to OpenAI for billions. Now that OpenAI can push whatever update or search to Chrome as default. Assuming they have use of it.

What would stop Google to build another browser say Information Explorer with the same engine and code? And market the hell out of it on its Web property?

> with the same engine and code

Presumably they would be selling the IP rights, so at least some kind of rewrite would be required, possibly without utilizing staff who worked on Chrome.

The Code is open source, they just get someone to package the browser into something different. Just like how M$ is doing with Edge?

It would be funny though if they hire different people to build another browser on top of webkit again XD.

...in rust, right? what an opportunity.
> What would stop Google to build another browser say Information Explorer with the same engine and code?

The courts. The courts would stop them. The entire premise for Google selling-off Chrome is a mandate that they divest themselves from the business itself.

> What would stop Google to build another browser say Information Explorer with the same engine and code? And market the hell out of it on its Web property?

Probably a court order, no? If you’re ordered to sell something, can you just recreate it immediately?

I dont know I have no idea. Could a court ruling bar certain entity from doing business in certain areas? Because That sounds silly to me. I wont be surprised if it was in other countries such as China, or EU , UK and Canada.

But US? The place that is perhaps the most pro Business or capitalistic on earth?

(comment deleted)
Well that would be a good way to kill off Chrome and get out of the mono-browser culture.
It is interesting how the data rush is changing, it was about Ads targeting, now it is about training AI. I wonder if one is better than the other for the open-source community? Would Chrome be more free or more locked-in with OpenAi
> it was about Ads targeting, now it is about training AI.

It's the same data rush. Don't fr a second think that "AI" will be used for anything but ads and selling your data t the highest bidder