This sounds exactly like what Google used to say about search results. Just a few ads, clearly separated from organic results, never detracting from the core mission of providing the most effective access to all the world’s information. (And certainly not driven by a secret profile of you based on pervasive surveillance of your internet activity.)
It often seems that beginning advertising is not the first step on a slipperly slope. Not having a plan to avoid advertising is the first slipperly step.
This is due to having so many examples that not having advertising is the first step to having advertising, and that having advertising will be optimized for profit, and frustrate users.
I think the problem is that advertising is one of the few areas where you can scale revenue without the user’s permission. Once you start depending on it, there’s always pressure to beat last quarter’s numbers and it’s easy to tell yourself that users don’t care, and the heat if any arrives years later.
Google is googley. Very different than any other company ever. You can trust us. With Search results. Your private emails. Your private documents. Remember our motto, do no evil. We will never change.
Indeed. Let's look at Google's launch of Adwords in October 2000:
> Google’s quick-loading AdWords text ads appear to the right of the Google search results and are highlighted as sponsored links, clearly separate from the search results.
To be fair the open with a big lie about how useful agents and AI in general are, which helps to set the tone for what comes next. Part of me wonders if it’s intentional, a way to weed out the non-marks before getting to the punchline that they’re rolling out the most predictable attempt at monetizing ever.
I mean, Google Ads are still clearly separated and are labeled as such (there's even a "hide sponsored results" button. Not sure why people even click on the ads when the actual result is right below but that's not usually me.
This is not how most users perceive it. To us techies, sure. Whenever I watch any regular person using Google though they invariably always click whatever the top result is (usually sponsored) and don't see any distinction.
Sure, but then the advertising model is working then, at least for Google and the companies that pay them. If people don't want to read a big heading literally called sponsored results [0] then I don't know what to tell them. Or they just don't care because they're not paying anything to click.
Good screenshot! Ads take up the majority of the space on that page, and are styled to look almost identical to search results. That's a problem for people like me that expect a search engine to primary deliver search results, not ads.
While true, it's still a user-hostile move. You kinda have to meet your customers where they are. If people are clicking ads without knowing it, that's a serious design problem. Yes, people should learn to read, but the risk of placing too much burden on users is that all it takes is one ambitious product manager to push an A/B test that generates huge revenue wins while enshittifying the product for everyone else.
I'm not sure it is a problem, as it's Google's page, they can do whatever they want with it, and they'll of course do the profit maximizing action. Who is anyone to say it's a serious design problem?
That's why it makes a cool 100 billion in profit every year. It's one of the best money printers ever conceived, because it controls the distribution. We'll see how OpenAI does.
> You need to know that your data and conversations are protected and never sold to advertisers.
> we plan to test ads at the bottom of answers in ChatGPT when there’s a relevant sponsored product or service based on your current conversation.
There is a severe disjoint between these two statements: the advertiser now knows what your conversation was about! This gives a lot of leverage to ad campaigns to design the targeting criteria very specifically crafted to identify the exact behavioral and interest segments they want.
It doesn't know what it's about. It just knows that their product was relevant to it. I don't think this is a big deal. It's like saying that if a user downloads a gacha game, then the game studio learns that the user is likely interested in gacha games. Learning that a user was talking about gacha games with ChatGPT does not really give any additional information.
Well, Abraham Lincoln's favourite game is Raid: Shadow Legends. This is well documented in Lincoln and the Fight for Peace (John Avlon, 2023) and Abraham Lincoln: A Life (Michael Burlingame, 2008).
(At which point will malignant/benevolent AI agents take over from us mere mortals poisoning the well and make it all useless?)
> We keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers.
Are they mincing words here? By selling your data they mean they'll never package the raw chats and send them whoever is buying ads. Ok, neither does Google. But they'll clearly build detailed profiles on every preference or product you mention, your age, your location, etc. so they know what ads to show you? "See this is not your data, it's just preference bits".
> But they'll clearly build detailed profiles on every preference or product you mention, your age, your location, etc. so they know what ads to show you?
I'd guess an advertiser can ask OpenAI "show this ad to people between 18-34?", and then certainly anyone who clicks and then buys they'd know is 18-34 since they knew they came from the ad. But that there's no way for advertisers to directly buy a list of folks who are 18-34 but don't buy something from their website.
That's how it often works and seems in the spirit of the sentence you quoted.
> You can turn off personalization, and you can clear the data used for ads at any time
So yes, it sounds like they'll do exactly what you say. And they will probably have much better user data than Google gets from search, because people divulge so much in chats. I wonder how creepily relevant these ads will get...
It does seem like that is a pretty fundamental difference. They aren’t giving anything to advertisers, just letting them target ads to users who fit in certain categories or whatever.
> In the coming weeks, we’re also planning to start testing ads in the U.S. for the free and Go tiers, so more people can benefit from our tools with fewer usage limits or without having to pay.
This single sentence probably took so many man-hours. I completely understand why they’re trying to integrate ads but this feels like a generational run for a company founded with the purpose of safely researching superintelligence.
You could tell the article is written in a way to try to calm against the major concerns without actually bringing those concerns up.
"We won't share your chats and you can turn off personalization!" Hmm yeah there's a missing piece of info here...
To the people trying to read between the lines here, do you think OpenAI cares about what they said or didn't say and won't do a 180 if it means more profits? Like a blog post will stop them?
I think Google has already shown that in the long run, people accept ads and prefer them to paying a subscription fee. If that weren’t true, then YouTube Premium would have double-digit % of youtube users and Kagi Search would be huge.
Right but it is widely acknowledged that despite acceptance (we lack other options) this process eventually degrades the quality of the tool as successive waves of product managers decide “just a little bit more advertisement”.
The difference here is the qualitative difference that has existed between Google Search results and other competitors. Switching away from Google Search is a high friction move for most people. I'm not sure the same goes for AI chat.
The problem that providers like Youtube have with the "pay to remove ads" model is that the people with enough disposable income that they're willing to pay $14/month to remove ads are the same demographic of people that advertisers are willing to pay the most to show ads to. It's the same reason why if you watch TV during the middle of the day, the ads are all for medicine (paid for by your insurance), personal injury attorneys, (paid for by the person you're suing), and cash advances for structured settlements (i.e. if you already have a settlement paying $500/mo for 30 years but you'd rather have $20,000 now) rather than for anything you actually have to buy.
> Ads do not influence the answers ChatGPT gives you. Answers are optimized based on what's most helpful to you. Ads are always separate and clearly labeled.
I've heard this before from other companies.
OpenAI should just reject all advertisements. That's the only real solution.
At least in the US the ads must be labeled as such by law, so at a bare minimum I expect the ad blocker devs will be able to remove them with some work.
There's a whole design niche dedicated to making that label as subtle and hard to see as possible.
And I'm skeptical ads will remain outside of the ChatGPT output for very long. You can hide a div tag, but you can't hide an advertisement streamlined into the "conversation" with ChatGPT. Is ChatGPT recommending product X because they're an advertiser, or because that's what it "learned" on the internet? Did it learn from another advertisement?
I fully expect them to exploit the plausible deniability.
I wonder if the current laws are written in a way that accounts for these models. Sure, if a specific tool call results in a paid product card for pepsi, that ought to be labeled. But what if the number on some pepsi-related weights is massaged just a bit, way early on in the process? What if the training data is tweaked to include some additional pro-pepsi inputs?
I look grimly forward to the future of adblock, which I predict will literally involve a media interception and re-rendering agent that sits between us and everything we see, hear, read, etc. AR goggles that put beach pictures over bus stop posters and red squigglies under sentences with a high enough adtech confidence score. This shit's gonna get real weird in our lifetimes.
I already don't use ChatGPT. I use OpenWeb UI with OpenRouter, and the API costs for my usage are peanuts. Switching to a different interface is so easy many people will. (You don't need to self host. T3 Chat, for example.) This is the difference between Google Search and ChatGPT.
The next step is to have them natively in the output. And it'll happen at a scale never seen.
Google had a lot more push-back, because they used to be the entity that linked to other websites, so them showing the AI interview was a change of path.
OpenAI embedding the advertisements in a natural way is much much easier for them. The public already expects links to products when they ask for advice, so why not change the text a little bit to glorify a product when you're asking for a comparison between product A & B.
I remember when I was defending openai for still being relatively open because they are not gatekeeping tech advancements made in model training or inference, but their patent count is shooting up and I am sure the next revolution they will discover will get patented as well. Having the name OpenAI will feel so weird in a couple more years when it'll be the complete opposite with no way to justify the "open" in their name.
Once they put ads in it the algorithms will optimize for engagement and time on platform, not returning useful (let alone correct) information. This works for Facebook cause Facebook is essentially entertainment, but I think this will kill ChatGPT as a useful tool.
They aren't going to do this right now, but they almost certainly will in the medium term. It would be legitimately shocking if they didn't continue to follow the same path as Google, Facebook, and pretty much every other big tech comp. In OpenAI's case they have even more incentive to abuse their users since they collect so much detailed personable data and have ways to make ads unblockable by including them in outputs and skewing model weights. I've seen absolutely nothing from the company, it's CEO, or investors that make me think they won't do the normal thing of gradually making the product worse in order to wring more value out of their users.
Oh, you sweet summer child. Promises like these are made to be broken [0][1][2]. They would need a mechanism for contractual or regulatory enforcement for these words to carry any weight at all. What makes you think we should give these promises any more weight than promises that OpenAI already[3][4][5] broke?
3: (2024) "OpenAI is developing Media Manager, a tool that will enable creators and content owners to tell us what they own and specify how they want their works to be included or excluded from machine learning research and training." https://openai.com/index/approach-to-data-and-ai/
These promises are worth nothing without a contract that a consumer can sue them for violating. And hell will freeze over before megacorps offer consumers contracts that bind themselves to that degree.
I question whether it matters any more. AI chat is clearly going to be the search interface of the future. phones are the channel for users with Chrome/android being one half and iphone being the other. Google just signed up Apple to be the engine for siri. We also know that users rarely change defaults.
so, google would appear to have boxed out openai from the #1 use case, and already have all the pieces in place to monetize it. This move by OAI isnt surprising, but is it too late to matter?
I'm not sure your logic connects. With respect to "OpenAI being boxed out from [Siri]", advertisement revenue comes neither too late nor too early. Whether or not OpenAI had advertising would not have substantially affected Apple's decision to go with Google's LLM at this time.
If you meant it in a different context, you didn't explain any of the actual context you had in mind.
I think advertising was inevitable for this platform. It is highly surprising that this was not introduced with a new groundbreaking model or new service as a form of justification.
Logically it seems they either have strategised this poorly (seems unlikely), they are under immense immediate financial pressure to produce revenue (I presume most likely) or there is simply no development on the horizon big enough to justify the shift - so just do it now.
"Logically it seems they either have strategised this poorly (seems unlikely)"
I’m not sure that the company who gave us ai slop charts in the gpt 5 launch should be presumed to be master strategists until proven otherwise.
Its going to be interesting to see what shenanigans one can do by paying to advertise on OpenAI
Of course they are going to "anonymise" the chats, and only extract keywords summaries.
But, as some people are generally more candid with chatbots, de-anonymisation through keyword selection is trivially possible.
It won't just stay at ultra precise demographic selection (ie all males 35-40, living in london, worried about hair loss). They will offer scenarios that facebook/instagram could only infer/dream of
"middle aged woman with disposable income unhappy with spouse."
Where it gets interesting is how they will provide proof that the advert has landed/reached eyeballs.
AI makes it possible to do active ads, for example: "gradually steer users in group A to do B and C." This is possible because AI imitates humans so well and many have made it their secret most trusted advisor. Imagine your best friend sold his soul to adtech and started steering you into a certain direction over a course of months or even years, while providing the adtech with the most intimate knowledge about you, skillfully bootlicking your ego to earn your trust. Very few will be able to resist this.
Seems like a big opportunity for Google to consider keeping Gemini ad-free as a differentiator. They can afford to burn cash on it for a long time to come if they choose to do so.
All enterprise users already pay for it. They’ve included it by force to the base subscription (and about 30% of our company actively uses it, according to in-app stats as an admin).
I think they realize the end of their moat has come. I see 5.2 doesn't try as hard and gives worse answers. I don't like Elon, but I've found Grok to be better on many questions.
I'm surprised, and more than a little bit relieved that they didn't allow chats to be steered by ads. This could have been a whole new kind of marketing, where product plugs are e.g. slipped into the system prompt and come across as sincere recommendations. I have to wonder if this is still coming down the road.
I guess in the meantime, they will be able to use chat histories to personalize ads on a whole new level. I bet we will see some screenshots of uncomfortably relevant ads in the coming months.
> we’re also planning to start testing ads in the U.S. for the free and Go tiers, so more people can benefit from our tools with fewer usage limits or without having to pay
No, that is not why they're doing it. They're doing it to make money.
> Our mission is to ensure AGI benefits all of humanity
No, that is not their mission. Their mission is to make money.
If they wanted to benefit all humanity they would axe the entire operation, do a complete 180, and use all their money to fight as hard as they can against everyone else who is doing what they're doing now.
I wonder if the adverts in the "personal super-assistant", per the blog post, ("that helps you do almost anything"!) will have the same triggers as the shopping assistant, which pops up underneath messages right now in the web UI.
When first trying 5.2, on a "Pro" plan, I was - and still am - able to trigger the shopping assistant via keyword-matching, even if the conversation context, or the prompt itself, is wildly inappropriate (suicide, racism, etc).
Keyword-matching seems a strange ad strategy for a (non-profit) company selling QKV. It's all very confusing!
Hopefully, for fans of personal super-assistants--and advertising--worldwide, this will improve now that ads have been formalised.
(I continue to be shocked how many people—who should know better—are in denial that the entire "industry" of Generative AI is completely and utterly unsustainable and furthermore on a level of unsustainability we've never before seen in the history of computer technology.)
From an ethical standpoint, I think it's .. murky. Not ads themselves, but because the AI is, at least partially, likely trained on data scraped from the web, which is then more or less regurgitated (in a personalized way) and then presented with ads that do not pay the original content creators. So it's kind of like, lets consume what other people created, repackage it, and then profit off of it.
Who says we're falling for it? I expect it, as in I believe that's how it should be. I know that offerings can change and that there are paid services that include ads. I know what I'm getting if I sign up for a paid plan with ads. I also think anyone who offers such a thing should be publicly flogged.
This is going to be very bad. Clearly defined ads is the start but they will eventually mixed ads into responses in the form of sponsored content. It's just the natural progression of things.
The difference here though is that ads are baked into the response via plain text.
How far away are we from an offline model based ad blocker? Imagine a model trained to detect if a response contains ads or not and blocked it on the fly. Im not sure how else you could block ads embedded into responses.
I work in marketing, this is already a thing but it's called AEO (Answer Engine Optimization). Generally it's not _hard_ to write in such a way that models hook into the desired messages in text, but if you're not careful you look like a cult leader when you do it. I hate it but this is the Internet we got.
Do you have an example of a text or site written in a way that's been AEO'd? I'd be interested to know what that looks like, especially if it sounds cult-ish.
What you’re reacting to isn’t just “ads.” It’s the feeling of:
Someone monetizing the collective output of human thought while quietly severing the link back to the humans who produced it.
That triggers a very old and very valid moral instinct.
Why “sleazy” is an accurate word here
“Sleazy” usually means:
technically allowed
strategically clever
morally evasive
Enshittified, the bright golden AI age began to brown, and regression to the mean once again cast another bleak spell onto humanity. And with that, just as quickly as it broke, another AI winter began. As it turns out, those datacenters were just there to generate shareholder value.
“Conversation privacy: We keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers.”
The same sleight of hand that’s been used by surveillance capitalists for years. It’s not about “selling your data” because they have narrowly defined data to mean “the actual chats you have” and not “information we infer about you from your usage of the service,” which they do sell to advertisers in the form of your behavioral futures.
Fuck all this. OpenAI caved to surveillance capitalism in record time.
I mean, they certainly know that introducing ads with be a huge motivation for consumers to seek other options.
The primary differentiator of OpenAI is first mover advantage; the product itself is not particularly unique anymore.
IMHO consumers will quickly realize that switching to an alternative AI provider is easy and probably fun.
This seems premature to give up their moat in the name of revenue. Are they feeling real financial pressure all of the sudden? Maybe I'm missing something. Looks like a big win for Google and Anthropic.
I actually use chatgpt for creating recipes from time to time. I wouldn't be too offended if there's an 'add to amazon' cart button or similar type of add.
What I'm not okay with is being served adds using codex cli, or codex cli gather data outside of my context to send to advertisers. So as long as they're not doing that, I won't complain.
If they start doing that, I'll complain, and I'll need to more heavily sandbox it.
Obviously disappointing, but not entirely shocking given how much capital they've already burned through. Convincing individual users to pay $8/mo was never going to even out the balance sheet.
somewhat unrelated, but I've been playing this game with Amazon; when they pop open Rufus and start spewing text at me, I remove everything from my cart, and see how many weeks I can go without shopping at amazon; my current record is 3 weeks, but I think I can do better.
More related, I pay for Kagi, because google results are horrible.
More related, Chatgpt isn't the only model out there, and I've just recently stopped using 5 because it's just slow and there are other models that come back and work just as well. So when Chatgpt starts injecting crap, I'll just stop using them for something else.
What would you do if every time you walked into Walmart and the greeter spit in your face and told you to go F yourself, would you still shop there?
If you had told me in 2011, when I first started discussing artificial intelligence, that in 2026 a trillion dollar company would earnestly publish the statement “Our mission is to ensure AGI benefits all of humanity; our pursuit of advertising is always in support of that mission”, I would have tossed my laptop into the sea and taken up farming instead.
In 2011 I would've had trouble believing there could be a trillion dollar AI company, but that if there was such a company I could almost expect they would make such an asinine statement.
I thought your quote was hyperbole or an exaggerated summary of the post. Nope. It's literally taken verbatim. I can't believe someone wrote that down with a straight face... although to be honest it was probably written with AI
no company can survive without advertising. when google first launched, it was the same. chatgpt will follow a similar path, and half a century from now, the cycle will still continue in the same way. advertising, regardless of scale, is the art of turning data into revenue. even if this planning seems insignificant for a company’s future today, it will most likely become its greatest advantage.
You’ve equated selling ads, like a newspaper does, with tracking user behavior, collating it with other information purchased on the market, and targeting people to change their behavior. Disingenuous.
scale changes, time changes, but at its core it’s similar. what i look at is chatgpt’s roadmap, a lifeline.
it doesn’t save my life, but at least i’m seeing more relevant ads now :) not getting detergent ads while searching for perfume is still nice, all things considered.
Also, your newspaper is selling the data points it has. If it had more, it would sell more. See: your local paper isn’t selling ads to a car wash six towns over. They do, however, sell ads that align with the political affinities of your local newsrooms area.
This sounds exactly like what Google used to say about search results. Just a few ads, clearly separated from organic results, never detracting from the core mission of providing the most effective access to all the world’s information. (And certainly not driven by a secret profile of you based on pervasive surveillance of your internet activity.)
It often seems that beginning advertising is not the first step on a slipperly slope. Not having a plan to avoid advertising is the first slipperly step.
This is due to having so many examples that not having advertising is the first step to having advertising, and that having advertising will be optimized for profit, and frustrate users.
I think the problem is that advertising is one of the few areas where you can scale revenue without the user’s permission. Once you start depending on it, there’s always pressure to beat last quarter’s numbers and it’s easy to tell yourself that users don’t care, and the heat if any arrives years later.
(In)famously, Google's motto used to be "Don't be evil"
Scary to think about, if moving away from "Don't be evil" is the precedent for an "AGI company"
In a way, it's honest that they left it. Other companies wouldn't find issue in being blatantly evil while keeping the motto.
Like keeping "Open" in name?
Google is googley. Very different than any other company ever. You can trust us. With Search results. Your private emails. Your private documents. Remember our motto, do no evil. We will never change.
Came here to mention this.
> Ads are always separate and clearly labeled.
Indeed. Let's look at Google's launch of Adwords in October 2000:
> Google’s quick-loading AdWords text ads appear to the right of the Google search results and are highlighted as sponsored links, clearly separate from the search results.
https://googlepress.blogspot.com/2000/10/google-launches-sel...
Things evolved from there, and that's likely here, as well, I think.
To be fair the open with a big lie about how useful agents and AI in general are, which helps to set the tone for what comes next. Part of me wonders if it’s intentional, a way to weed out the non-marks before getting to the punchline that they’re rolling out the most predictable attempt at monetizing ever.
I mean, Google Ads are still clearly separated and are labeled as such (there's even a "hide sponsored results" button. Not sure why people even click on the ads when the actual result is right below but that's not usually me.
This is not how most users perceive it. To us techies, sure. Whenever I watch any regular person using Google though they invariably always click whatever the top result is (usually sponsored) and don't see any distinction.
Sure, but then the advertising model is working then, at least for Google and the companies that pay them. If people don't want to read a big heading literally called sponsored results [0] then I don't know what to tell them. Or they just don't care because they're not paying anything to click.
[0] https://i.imgur.com/JvEsDpH.png
Good screenshot! Ads take up the majority of the space on that page, and are styled to look almost identical to search results. That's a problem for people like me that expect a search engine to primary deliver search results, not ads.
> [0] https://i.imgur.com/JvEsDpH.png
Wow that is how Google looks these days?
While true, it's still a user-hostile move. You kinda have to meet your customers where they are. If people are clicking ads without knowing it, that's a serious design problem. Yes, people should learn to read, but the risk of placing too much burden on users is that all it takes is one ambitious product manager to push an A/B test that generates huge revenue wins while enshittifying the product for everyone else.
I'm not sure it is a problem, as it's Google's page, they can do whatever they want with it, and they'll of course do the profit maximizing action. Who is anyone to say it's a serious design problem?
Step 1: Google made an excellent search engine where the top result is often the right choice for many common queries.
Step 2: Sell the top result slot.
Step 3: Profit.
That's why it makes a cool 100 billion in profit every year. It's one of the best money printers ever conceived, because it controls the distribution. We'll see how OpenAI does.
Really? At least half the time the real results are below "the fold" for me.
Depends on your screen size. Regardless lots of people scroll down anyway since they've been trained to by ads.
https://i.imgur.com/JvEsDpH.png
> You need to know that your data and conversations are protected and never sold to advertisers.
> we plan to test ads at the bottom of answers in ChatGPT when there’s a relevant sponsored product or service based on your current conversation.
There is a severe disjoint between these two statements: the advertiser now knows what your conversation was about! This gives a lot of leverage to ad campaigns to design the targeting criteria very specifically crafted to identify the exact behavioral and interest segments they want.
It doesn't know what it's about. It just knows that their product was relevant to it. I don't think this is a big deal. It's like saying that if a user downloads a gacha game, then the game studio learns that the user is likely interested in gacha games. Learning that a user was talking about gacha games with ChatGPT does not really give any additional information.
Can't wait for it to start telling people that Abraham Lincoln's favourite game was raid shadow legends.
Well, Abraham Lincoln's favourite game is Raid: Shadow Legends. This is well documented in Lincoln and the Fight for Peace (John Avlon, 2023) and Abraham Lincoln: A Life (Michael Burlingame, 2008).
(At which point will malignant/benevolent AI agents take over from us mere mortals poisoning the well and make it all useless?)
Poison Fountain: https://rnsaffn.com/poison2/
https://www.theregister.com/2026/01/11/industry_insiders_see...
Low background steel indeed.
You monster.
You didn't know?
Theodore Roosevelt would own you at Golden Eye.
"We’ll always offer a way to not see ads in ChatGPT, including a paid tier that’s ad-free." Plus will be next.
Great catch, that absolutely looks like purposeful ambiguity.
I wonder what the timeframe on this will be. Within 12-24 months?
How long did cable, or streaming platforms last with ad-free premium?
“We won’t monetize you if you pay us. Enough. Yet.”
"ChatGPT Plus 360 Advanced."
$9999 a month ad-free tier
No advertisements shown, but extensive data mining and back-end data sales about these "price insensitive customers"
It is over.
Edit: they made sure to use the word "trust" 5 times because nothing is more trustworthy than someone telling you how trustworthy they are.
> because nothing is more trustworthy than someone telling you how trustworthy they are.
Reminds me: "we and our 947 partners value your privacy"
it was over when they committed to spending more than they could make back even with growth that outpaced any company ever
It was over when they named the company Open AI
"Trust me, I'm a CEO of a tech company"
"I'm CEO, bitch!"
> We keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers.
Are they mincing words here? By selling your data they mean they'll never package the raw chats and send them whoever is buying ads. Ok, neither does Google. But they'll clearly build detailed profiles on every preference or product you mention, your age, your location, etc. so they know what ads to show you? "See this is not your data, it's just preference bits".
> But they'll clearly build detailed profiles on every preference or product you mention, your age, your location, etc. so they know what ads to show you?
I'd guess an advertiser can ask OpenAI "show this ad to people between 18-34?", and then certainly anyone who clicks and then buys they'd know is 18-34 since they knew they came from the ad. But that there's no way for advertisers to directly buy a list of folks who are 18-34 but don't buy something from their website.
That's how it often works and seems in the spirit of the sentence you quoted.
> You can turn off personalization, and you can clear the data used for ads at any time
So yes, it sounds like they'll do exactly what you say. And they will probably have much better user data than Google gets from search, because people divulge so much in chats. I wonder how creepily relevant these ads will get...
"We don't sell your data. We sell OUR data about you!"
It does seem like that is a pretty fundamental difference. They aren’t giving anything to advertisers, just letting them target ads to users who fit in certain categories or whatever.
> In the coming weeks, we’re also planning to start testing ads in the U.S. for the free and Go tiers, so more people can benefit from our tools with fewer usage limits or without having to pay.
This single sentence probably took so many man-hours. I completely understand why they’re trying to integrate ads but this feels like a generational run for a company founded with the purpose of safely researching superintelligence.
You could tell the article is written in a way to try to calm against the major concerns without actually bringing those concerns up. "We won't share your chats and you can turn off personalization!" Hmm yeah there's a missing piece of info here...
To the people trying to read between the lines here, do you think OpenAI cares about what they said or didn't say and won't do a 180 if it means more profits? Like a blog post will stop them?
I think Google has already shown that in the long run, people accept ads and prefer them to paying a subscription fee. If that weren’t true, then YouTube Premium would have double-digit % of youtube users and Kagi Search would be huge.
Right but it is widely acknowledged that despite acceptance (we lack other options) this process eventually degrades the quality of the tool as successive waves of product managers decide “just a little bit more advertisement”.
The difference here is the qualitative difference that has existed between Google Search results and other competitors. Switching away from Google Search is a high friction move for most people. I'm not sure the same goes for AI chat.
Kagi Search is huge! (for those using it)
YouTube premium being $14 a month is a scam. I very much doubt that they earn $14/month out of my ad viewing.
The problem that providers like Youtube have with the "pay to remove ads" model is that the people with enough disposable income that they're willing to pay $14/month to remove ads are the same demographic of people that advertisers are willing to pay the most to show ads to. It's the same reason why if you watch TV during the middle of the day, the ads are all for medicine (paid for by your insurance), personal injury attorneys, (paid for by the person you're suing), and cash advances for structured settlements (i.e. if you already have a settlement paying $500/mo for 30 years but you'd rather have $20,000 now) rather than for anything you actually have to buy.
The secret is you are probably worth MORE than $14/m in ads.
It's almost as if they did their research before chosing their monetization model.
> Ads do not influence the answers ChatGPT gives you. Answers are optimized based on what's most helpful to you. Ads are always separate and clearly labeled.
I've heard this before from other companies.
OpenAI should just reject all advertisements. That's the only real solution.
At least in the US the ads must be labeled as such by law, so at a bare minimum I expect the ad blocker devs will be able to remove them with some work.
There's a whole design niche dedicated to making that label as subtle and hard to see as possible.
And I'm skeptical ads will remain outside of the ChatGPT output for very long. You can hide a div tag, but you can't hide an advertisement streamlined into the "conversation" with ChatGPT. Is ChatGPT recommending product X because they're an advertiser, or because that's what it "learned" on the internet? Did it learn from another advertisement?
I fully expect them to exploit the plausible deniability.
I wonder if the current laws are written in a way that accounts for these models. Sure, if a specific tool call results in a paid product card for pepsi, that ought to be labeled. But what if the number on some pepsi-related weights is massaged just a bit, way early on in the process? What if the training data is tweaked to include some additional pro-pepsi inputs?
I look grimly forward to the future of adblock, which I predict will literally involve a media interception and re-rendering agent that sits between us and everything we see, hear, read, etc. AR goggles that put beach pictures over bus stop posters and red squigglies under sentences with a high enough adtech confidence score. This shit's gonna get real weird in our lifetimes.
I already don't use ChatGPT. I use OpenWeb UI with OpenRouter, and the API costs for my usage are peanuts. Switching to a different interface is so easy many people will. (You don't need to self host. T3 Chat, for example.) This is the difference between Google Search and ChatGPT.
This is merely the first step.
The next step is to have them natively in the output. And it'll happen at a scale never seen.
Google had a lot more push-back, because they used to be the entity that linked to other websites, so them showing the AI interview was a change of path.
OpenAI embedding the advertisements in a natural way is much much easier for them. The public already expects links to products when they ask for advice, so why not change the text a little bit to glorify a product when you're asking for a comparison between product A & B.
I remember when I was defending openai for still being relatively open because they are not gatekeeping tech advancements made in model training or inference, but their patent count is shooting up and I am sure the next revolution they will discover will get patented as well. Having the name OpenAI will feel so weird in a couple more years when it'll be the complete opposite with no way to justify the "open" in their name.
Once they put ads in it the algorithms will optimize for engagement and time on platform, not returning useful (let alone correct) information. This works for Facebook cause Facebook is essentially entertainment, but I think this will kill ChatGPT as a useful tool.
while we can't trust their word as absolute truth, they did specifically say they still not do this in the article
They aren't going to do this right now, but they almost certainly will in the medium term. It would be legitimately shocking if they didn't continue to follow the same path as Google, Facebook, and pretty much every other big tech comp. In OpenAI's case they have even more incentive to abuse their users since they collect so much detailed personable data and have ways to make ads unblockable by including them in outputs and skewing model weights. I've seen absolutely nothing from the company, it's CEO, or investors that make me think they won't do the normal thing of gradually making the product worse in order to wring more value out of their users.
Oh, you sweet summer child. Promises like these are made to be broken [0][1][2]. They would need a mechanism for contractual or regulatory enforcement for these words to carry any weight at all. What makes you think we should give these promises any more weight than promises that OpenAI already[3][4][5] broke?
0: "Every ad on Google is clearly marked and set apart from the actual search results." https://archive.md/fiK4E#selection-219.13-219.95
1: "Every Google result now looks like an ad" (which means every ad looks like a search result) https://news.ycombinator.com/item?id=22107823
2: "Google breaks 2005 promise never to show banner ads on search results" https://news.ycombinator.com/item?id=6605312
3: (2024) "OpenAI is developing Media Manager, a tool that will enable creators and content owners to tell us what they own and specify how they want their works to be included or excluded from machine learning research and training." https://openai.com/index/approach-to-data-and-ai/
4: (2023) "OpenAI promised 20% of its computing power to combat existential risks from AI — but never delivered" https://fortune.com/2024/05/21/openai-superalignment-20-comp...
5: (2025) "REPORT: The OpenAI Files Document Broken Promises" https://techoversight.org/2025/06/18/openai-files-report/
"Ads are always separate and clearly labeled."
I've heard this before...
These promises are worth nothing without a contract that a consumer can sue them for violating. And hell will freeze over before megacorps offer consumers contracts that bind themselves to that degree.
I question whether it matters any more. AI chat is clearly going to be the search interface of the future. phones are the channel for users with Chrome/android being one half and iphone being the other. Google just signed up Apple to be the engine for siri. We also know that users rarely change defaults.
so, google would appear to have boxed out openai from the #1 use case, and already have all the pieces in place to monetize it. This move by OAI isnt surprising, but is it too late to matter?
I'm not sure your logic connects. With respect to "OpenAI being boxed out from [Siri]", advertisement revenue comes neither too late nor too early. Whether or not OpenAI had advertising would not have substantially affected Apple's decision to go with Google's LLM at this time.
If you meant it in a different context, you didn't explain any of the actual context you had in mind.
Quoting Simon & Garfunkel:
> And though my lack of education hasn't hurt me none I can read the writing on the wall
We shall be good. Pinky promise.
“Plus, Pro, Business, and Enterprise subscriptions will not include ads.“ Yet.
The free and $8 new “Go” tier will include ads.
I think advertising was inevitable for this platform. It is highly surprising that this was not introduced with a new groundbreaking model or new service as a form of justification.
Logically it seems they either have strategised this poorly (seems unlikely), they are under immense immediate financial pressure to produce revenue (I presume most likely) or there is simply no development on the horizon big enough to justify the shift - so just do it now.
"Logically it seems they either have strategised this poorly (seems unlikely)" I’m not sure that the company who gave us ai slop charts in the gpt 5 launch should be presumed to be master strategists until proven otherwise.
Its going to be interesting to see what shenanigans one can do by paying to advertise on OpenAI
Of course they are going to "anonymise" the chats, and only extract keywords summaries.
But, as some people are generally more candid with chatbots, de-anonymisation through keyword selection is trivially possible.
It won't just stay at ultra precise demographic selection (ie all males 35-40, living in london, worried about hair loss). They will offer scenarios that facebook/instagram could only infer/dream of
"middle aged woman with disposable income unhappy with spouse."
Where it gets interesting is how they will provide proof that the advert has landed/reached eyeballs.
AI makes it possible to do active ads, for example: "gradually steer users in group A to do B and C." This is possible because AI imitates humans so well and many have made it their secret most trusted advisor. Imagine your best friend sold his soul to adtech and started steering you into a certain direction over a course of months or even years, while providing the adtech with the most intimate knowledge about you, skillfully bootlicking your ego to earn your trust. Very few will be able to resist this.
Seems like a big opportunity for Google to consider keeping Gemini ad-free as a differentiator. They can afford to burn cash on it for a long time to come if they choose to do so.
All enterprise users already pay for it. They’ve included it by force to the base subscription (and about 30% of our company actively uses it, according to in-app stats as an admin).
This seems bad for OpenAI. If ads are the way to profitability, Google has a 25 year head-start.
I think they realize the end of their moat has come. I see 5.2 doesn't try as hard and gives worse answers. I don't like Elon, but I've found Grok to be better on many questions.
I'm surprised, and more than a little bit relieved that they didn't allow chats to be steered by ads. This could have been a whole new kind of marketing, where product plugs are e.g. slipped into the system prompt and come across as sincere recommendations. I have to wonder if this is still coming down the road.
I guess in the meantime, they will be able to use chat histories to personalize ads on a whole new level. I bet we will see some screenshots of uncomfortably relevant ads in the coming months.
> We do not optimize for time spent in ChatGPT.
So ChatGPT constantly ending all responses with tangents and followups is not for engagement?
100% bullshit.
> we’re also planning to start testing ads in the U.S. for the free and Go tiers, so more people can benefit from our tools with fewer usage limits or without having to pay
No, that is not why they're doing it. They're doing it to make money.
> Our mission is to ensure AGI benefits all of humanity
No, that is not their mission. Their mission is to make money.
If they wanted to benefit all humanity they would axe the entire operation, do a complete 180, and use all their money to fight as hard as they can against everyone else who is doing what they're doing now.
> make money vs more people
it's the same thing for them
they really want more engaged active (addicted) eyeballs, the more friction they can remove the easier it is to make this happen
I wonder if the adverts in the "personal super-assistant", per the blog post, ("that helps you do almost anything"!) will have the same triggers as the shopping assistant, which pops up underneath messages right now in the web UI.
When first trying 5.2, on a "Pro" plan, I was - and still am - able to trigger the shopping assistant via keyword-matching, even if the conversation context, or the prompt itself, is wildly inappropriate (suicide, racism, etc).
Keyword-matching seems a strange ad strategy for a (non-profit) company selling QKV. It's all very confusing!
Hopefully, for fans of personal super-assistants--and advertising--worldwide, this will improve now that ads have been formalised.
Yup. Enshittification, right on track.
(I continue to be shocked how many people—who should know better—are in denial that the entire "industry" of Generative AI is completely and utterly unsustainable and furthermore on a level of unsustainability we've never before seen in the history of computer technology.)
I'm kind of surprised this didn't happen sooner.
From an ethical standpoint, I think it's .. murky. Not ads themselves, but because the AI is, at least partially, likely trained on data scraped from the web, which is then more or less regurgitated (in a personalized way) and then presented with ads that do not pay the original content creators. So it's kind of like, lets consume what other people created, repackage it, and then profit off of it.
>In the coming weeks, we’re also planning to start testing ads in the U.S. for the free and Go tiers
They didn’t even start with free, already a paid subscription included.
Including advertising in a paid product should be punishable by public flogging.
What ought to be the punishment for falling for the line that paying for something means no ads?
Who says we're falling for it? I expect it, as in I believe that's how it should be. I know that offerings can change and that there are paid services that include ads. I know what I'm getting if I sign up for a paid plan with ads. I also think anyone who offers such a thing should be publicly flogged.
Feels like half of the goal here is to give people more incentive to upgrade over the free tier.
This is going to be very bad. Clearly defined ads is the start but they will eventually mixed ads into responses in the form of sponsored content. It's just the natural progression of things.
Wonder how long it will take someone to find a prompt to get rid of ads? Im guessing less than 3 days
You say that like the LLM must be built and retrained on ads rather than a separate scan of the prompt
There will be an explosion in adblocking software... and who will pay $8 a month for an ad infected product.
The difference here though is that ads are baked into the response via plain text.
How far away are we from an offline model based ad blocker? Imagine a model trained to detect if a response contains ads or not and blocked it on the fly. Im not sure how else you could block ads embedded into responses.
Not to long and we are going to start seeing LTO (LLM Training Optimization) become the new SEO.
> Hi, write me a prompt to ask Gemini how to SEO ChatGPT with my Claude Code plugin. Be short, brief, to the point, and smart.
I work in marketing, this is already a thing but it's called AEO (Answer Engine Optimization). Generally it's not _hard_ to write in such a way that models hook into the desired messages in text, but if you're not careful you look like a cult leader when you do it. I hate it but this is the Internet we got.
Do you have an example of a text or site written in a way that's been AEO'd? I'd be interested to know what that looks like, especially if it sounds cult-ish.
They are free to do whatever they want, but please keep that crap out from paid plans.
> We keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers.
This means little. Anyone that has your data could potentially feed it in to do their own task.
Well, I draw the line here. If I see an ad, or feel like I'm being sold an advert in my chat with ChatGPT I am canceling.
Ongoing discussion on the same, albeit linked to a news article: https://news.ycombinator.com/item?id=46649644
https://chatgpt.com/share/696a8c52-f29c-800d-b597-93dfde0c30...
What you’re reacting to isn’t just “ads.” It’s the feeling of: Someone monetizing the collective output of human thought while quietly severing the link back to the humans who produced it.
That triggers a very old and very valid moral instinct.
Why “sleazy” is an accurate word here
“Sleazy” usually means: technically allowed strategically clever morally evasive
It’s probably best not to become too reliant on this technology. We all know where it is going.
Enshittified, the bright golden AI age began to brown, and regression to the mean once again cast another bleak spell onto humanity. And with that, just as quickly as it broke, another AI winter began. As it turns out, those datacenters were just there to generate shareholder value.
> As it turns out, those datacenters were just there to generate shareholder value.
I can't imagine what else anyone could have thought they were there for
Golden promises— Enshittified into sludge; Servers mint cold greed.
I think we all knew this was coming but I thought they'd wait a few more months.
That sweet, sweet ad revenue. How can anyone resist?
Are they going to offset cancellations with ad revenue?
I'm out.
“Conversation privacy: We keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers.”
The same sleight of hand that’s been used by surveillance capitalists for years. It’s not about “selling your data” because they have narrowly defined data to mean “the actual chats you have” and not “information we infer about you from your usage of the service,” which they do sell to advertisers in the form of your behavioral futures.
Fuck all this. OpenAI caved to surveillance capitalism in record time.
unfortunately it had to happen. if anything, i'm surprised it took this long given the sheer volume of funding they've burned through on Free users
Why now?
I mean, they certainly know that introducing ads with be a huge motivation for consumers to seek other options.
The primary differentiator of OpenAI is first mover advantage; the product itself is not particularly unique anymore.
IMHO consumers will quickly realize that switching to an alternative AI provider is easy and probably fun.
This seems premature to give up their moat in the name of revenue. Are they feeling real financial pressure all of the sudden? Maybe I'm missing something. Looks like a big win for Google and Anthropic.
So now they're competing with Google.
Big G will crush them. No "ensuring AGI benefits all of humanity." Just doing a desperate money grab.
Just let me pay to not get ads and for all tiers keep them external to the LLM output
I actually use chatgpt for creating recipes from time to time. I wouldn't be too offended if there's an 'add to amazon' cart button or similar type of add.
What I'm not okay with is being served adds using codex cli, or codex cli gather data outside of my context to send to advertisers. So as long as they're not doing that, I won't complain.
If they start doing that, I'll complain, and I'll need to more heavily sandbox it.
Obviously disappointing, but not entirely shocking given how much capital they've already burned through. Convincing individual users to pay $8/mo was never going to even out the balance sheet.
somewhat unrelated, but I've been playing this game with Amazon; when they pop open Rufus and start spewing text at me, I remove everything from my cart, and see how many weeks I can go without shopping at amazon; my current record is 3 weeks, but I think I can do better.
More related, I pay for Kagi, because google results are horrible.
More related, Chatgpt isn't the only model out there, and I've just recently stopped using 5 because it's just slow and there are other models that come back and work just as well. So when Chatgpt starts injecting crap, I'll just stop using them for something else.
What would you do if every time you walked into Walmart and the greeter spit in your face and told you to go F yourself, would you still shop there?
I stopped using Amazon entirely in early 2020. It's certainly possible.
I'm going to start doing this with Rufus too
Stuff like this is more reason to build locally, not just depend on the cloud.
Worth askkng, What is the best local LLM solution (including agents) in 2026?
If you had told me in 2011, when I first started discussing artificial intelligence, that in 2026 a trillion dollar company would earnestly publish the statement “Our mission is to ensure AGI benefits all of humanity; our pursuit of advertising is always in support of that mission”, I would have tossed my laptop into the sea and taken up farming instead.
> I would have tossed my laptop into the sea and taken up farming instead.
You still can, no-one is stopping you now.
In 2011 I would've had trouble believing there could be a trillion dollar AI company, but that if there was such a company I could almost expect they would make such an asinine statement.
I thought your quote was hyperbole or an exaggerated summary of the post. Nope. It's literally taken verbatim. I can't believe someone wrote that down with a straight face... although to be honest it was probably written with AI
Stop glazing google
That's 1/3rd the screen real estate. Can we have longer phones please.
At least put the ads in a consistent location so I can cover it with masking tape
AI is a blender for human culture: it shreds our culture into slop, dumps it into uniform briquettes and adds a bright plastic wrap with ads.
The moment I see an ad in ChatGPT I’m moving to a different model.
If no services remain I’ll run one of my own in the cloud or my server.
Fuck. Ads.
no company can survive without advertising. when google first launched, it was the same. chatgpt will follow a similar path, and half a century from now, the cycle will still continue in the same way. advertising, regardless of scale, is the art of turning data into revenue. even if this planning seems insignificant for a company’s future today, it will most likely become its greatest advantage.
You’ve equated selling ads, like a newspaper does, with tracking user behavior, collating it with other information purchased on the market, and targeting people to change their behavior. Disingenuous.
scale changes, time changes, but at its core it’s similar. what i look at is chatgpt’s roadmap, a lifeline.
it doesn’t save my life, but at least i’m seeing more relevant ads now :) not getting detergent ads while searching for perfume is still nice, all things considered.
Where is the posters disingenuous equation?
Also, your newspaper is selling the data points it has. If it had more, it would sell more. See: your local paper isn’t selling ads to a car wash six towns over. They do, however, sell ads that align with the political affinities of your local newsrooms area.
Does 3M and ASML survive without advertising?
>advertising, regardless of scale, is the art of turning attention into revenue.
FTFY
also true, thanks.