ChatGPT is getting ads soon, starting with free users

[Daniel Sims]

What just happened? The day many ChatGPT users have feared for weeks has finally arrived. The chatbot that has become nearly synonymous with generative AI will soon begin testing advertising. Ads will appear for some free users as well as subscribers on the service's newest and least expensive tier, which became available worldwide this week.

ChatGPT users in the US will start seeing ads in the coming weeks unless they subscribe to the Plus or Pro tiers. The chatbot will not show ads to users who indicate they are under 18 or if it detects that a user is a minor.

Ads will generally appear below the AI's responses. While the ads will be contextually relevant to the conversation, OpenAI emphasized that they will not influence the chatbot's answers and that advertisers will not have access to user conversations.

For example, one screenshot shows a user asking for dinner party ideas, with an ad for hot sauce – clearly marked as sponsored – appearing below ChatGPT's recipe suggestions. In another screenshot, an answer about Santa Fe is followed by an ad from a business offering to help plan a trip there. Tapping a button below the ad opens a second chatbot window where users can request more information.

ChatGPT will not display ads during conversations about health or politics. Users can also turn off ad personalization and delete the data used to tailor ads.

Indications that OpenAI planned to introduce ads first appeared in October, when The Information reported that the company had hired hundreds of former Meta employees. A developer later discovered code referencing ads in a beta version of ChatGPT's Android app.

In related news, US users can now subscribe to ChatGPT's new Go tier, which OpenAI began testing last year in India and other countries. Priced at $8 per month, the plan offers 10 times more interactions than the free tier, access to the latest GPT-5.2 model, and an expanded memory.

It is unsurprising that OpenAI is seeking ad revenue and encouraging more subscriptions. The company has not yet reached profitability, and roughly 95 percent of its nearly one billion weekly users still rely on the free tier.

ChatGPT also quietly expanded into machine translation this week. The new tool currently mirrors Google Translate's web interface, supports 47 languages, and allows switching between four different tones. Although options for image and audio translation are visible, these features are not yet functional.

Permalink to story:

ChatGPT is getting ads soon, starting with free users

 

[Haaky]

Not a suprise, nothing in this world is free, not even Freedom

 

[JMGuy]

And nothing of actual value was lost. ChatGPT is ok for some super basic things—like quick document editing and summing up a page of text and such—but that isn’t something I find useful enough to pay for. Overall the output is typically pretty poor and requires a lot of editing anyway.

 

[yRaz]

A service adding Ads has always been a redflag that it's in trouble

 

[Hexic]

They’re even going to show ads to paid subscribers? Yeah, screw that. Glad I don’t use GPT.

 

[MasterAce]

We are at the top of a slippery slope. Adverisers would trade their mom to get that data, and openai happens to be desperate for some cash. In their minds they can’t normalize adds and tracking fast enough.

Give it a year and they are selling our data like candy, I’m certain of it.

 

[Anton Longshot]

If the Plan is to pay off their debts off their ad-pushing income I have bad news for Altman & Co : The end of the Universe will happen before that debt is paid.

 

[Theinsanegamer]

And watch as consumers LEAVE since there is competition that doesnt have the ads. Especially the PAID users, AKA your only chance at making a profit.

 

[AlexIL]

Its a test how people react. If it goes well you’ll see ads in every AI.

 

[JMGuy]

A service adding Ads has always been a redflag that it's in trouble

Considering how many AI-space CEOs have recently gone on the offensive complaining their amazing tech is just not getting enough love, in conjunction with this move, it does make one wonder if that bubble is starting to make a few people sweat.

 

[JMGuy]

It’s a test how people react. If it goes well you’ll see ads in every AI.

I really think it’s just too soon. Like yRaz suggested, this looks like a strong indicator of trouble. Moving to ads this early suggests subscriptions alone aren’t covering costs, and that investors are already pushing for returns to cut losses. When the bubble bursts, it’s increasingly likely that Google and Microsoft will be left standing with few real competitors—their AI costs are subsidized by entrenched service ecosystems. ChatGPT and others, without that kind of fallback, look weaker by the day.

 

[godrilla]

Considering how many AI-space CEOs have recently gone on the offensive complaining they’re amazing tech is just not getting enough love, in conjunction with this move, it does make one wonder if that bubble is starting to make a few people sweat.

investors are starting to catch on.
Oracle getting sued over ai investment.
www.tomshardware.com https://share.google/Fm9mFyLfEGbAIE4yn

 

[yRaz]

Considering how many AI-space CEOs have recently gone on the offensive complaining their amazing tech is just not getting enough love, in conjunction with this move, it does make one wonder if that bubble is starting to make a few people sweat.

The only redflag I need is that the massive foreign corporations producing the hardware for American AI infrastructure do not see it worthwhile to invest those profits in themselves to expand manufacturing capacity for "demand".

Further, I see AI being most useful when it is trained on VERY SPECIFIC problems, like AlphaFold. We will not see an AGI any time soon and pretending that we can turn LLMs into an AGI by just throwing money at the problem is absurd

 

[Neatfeatguy]

So a company that is given billions of dollars from investors needs ad revenue to help keep the lights on.....isn't that a red flag to others?

 

[Watzupken]

In my opinion, doing advertisement is one way to make money. However,
1. The ad "pie" is limited, so it cannot save all the AI companies out there, and,

2. The likes of OpenAi is going to face an uphill challenge against Google, that's got a very strong foothold in ads.

Ultimately, there is nothing they can do to fully cover their unsustainable investment. Once investor stops throwing money into a bottomless pit and banks stop lending due to increasing risk to lend to such industry, the bubble will burst.

 

[James321]

Have anyone ever heard another person say they going to buy a product because they saw an ad?

 

[godrilla]

Have anyone ever heard another person say they going to buy a product because they saw an ad?

Ubisoft spent 350 million USD on Starwars Outlaws and that game is considered a failure. There is a significant disconnect between the expenditure on ads/marketing and the return on the investment. In 2026 imo a great product will sell itself by word of mouth vs paid influencers and marketing.
I wonder when/if the bubble burts and it's time for bailouts who will survive the storn now that Nvidia has thier own LLM via Groq?

 

[Vanderlinde]

Have anyone ever heard another person say they going to buy a product because they saw an ad?

No, but stink into links or content provided by AI I did see them yes. Or even in my own case, a recommendation program to do something locally, suggested by AI, which came with a BOATLOAD of bloatware without even giving me the option to agree or disagree. It took me 15 minutes to deinstall all that.

We will have the same stuff going on as with Google, Meta and such in the past and now, where ads will be used for scam, false information and what more.

The free service is just another tool for them to train masses of data provided by millions of interactions on a daily basis. I think the data at this point is getting stuffed.

Any service like google, apple, meta and such free AI means it trains on your input.

 

[ScottSoapbox]

Google turned free search into billions in ads.

AI is already replacing search so ads is an obvious monetizing for free users.

The true test is how quickly it enshitifies. Google took many years, but AI is burning cash at unprecedented rates.

 

[VitalyT]

I said this was going to happen shortly after the free-cheese rollout, that they were going to stick it in. They were just posturing, while looking around for a lubricant.

 

[AlexIL]

I really think it’s just too soon. Like yRaz suggested, this looks like a strong indicator of trouble. Moving to ads this early suggests subscriptions alone aren’t covering costs, and that investors are already pushing for returns to cut losses. When the bubble bursts, it’s increasingly likely that Google and Microsoft will be left standing with few real competitors—their AI costs are subsidized by entrenched service ecosystems. ChatGPT and others, without that kind of fallback, look weaker by the day.

ChatGPT fell behind Gemini in quality quite fast. OpenAI won’t survive the market as is unless Microsoft absorbs it but by now it’s just dead weight for the corp. Altman should have thought twice before doing his moves, his CSO warned him but greed, well now he deals with his own demise. I pulled the plug on ChatGPT already.

 

[OortCloud]

What do you do with a technology that is very expensive to set up and run and doesn't work and is being found to perform very poorly in the real world and is very expensive for businesses to use and the economic model makes no sense for non-paying consumers? Add some adverts - that will make it much better. Advertised Incompetence for the win.

 

[pnntmp]

A service adding Ads has always been a redflag that it's in trouble

This is Nonsense with a capital N.

You're saying Google, Meta and everything else that relies on ads as a main revenue source are "in trouble"? Seriously?

 

[yRaz]

This is Nonsense with a capital N.

You're saying Google, Meta and everything else that relies on ads as a main revenue source are "in trouble"? Seriously?

most of the internet runs off of google and meta's ad services and data collection. using them as an example is like pointing out how car dealerships sell cars

 

[Axeia]

So...
May 2024: I kind of think of ads as a last resort for us for a business model
June 2025: We haven't done any advertising product yet. I kind of...I mean, I'm not totally against it
Jan 2026: this article announcing ads, even in the cheapest paid tier.

At least Sam Altman never claimed they wouldn't have them I suppose.

Now in my opinion I don't think they'll help all that much if they're like in the examples above. Revenue from simple text ads is extremely low, costs of AI interactions are quite high.
Next step, make the ads blend in more to get them accidental clicks.
Step after that, full screen take overs with a countdown until you can skip them.
Step after that, have the AI steer the conversation into the direction of purchasing something by custom tailoring them to your psychological profile. "You know what really would help you with this task to? An Adobe subscription. The majority of people in your sector use it to great success and you risk falling behind if you don't".

And it won't just be OpenAI either. This will go exactly like it did with subscription streaming services. Netflix is the big boy they're and led the charge with the rest following. Here it'll be OpenAI first with the rest following. The only thing I'm surprised by that this isn't as much directly driven by greed from the company but most likely investors finally demanding to see some kind of actual business model.