Coming soon: We interrupt this ChatGPT session with a very special message from our sponsors
(2026/01/17)
- Reference: 1768610472
- News link: https://www.theregister.co.uk/2026/01/17/openai_chatgpt_ads/
- Source link:
OpenAI's budget ChatGPT Go subscription tier has migrated to the US, soon to be accompanied by advertising. The company's free tier will be similarly afflicted.
ChatGPT Go, the company's cheapest paid offering, has been an option in 171 countries since August and is now available stateside for $8 per month.
"In the coming weeks, we're also planning to start testing ads in the US for the free and Go tiers, so more people can benefit from our tools with fewer usage limits or without having to pay," [1]said Fidji Simo, OpenAI's CEO of applications. "Plus, Pro, Business, and Enterprise subscriptions will not include ads."
[2]
About 5 percent, or 35 million, of ChatGPT's weekly active user base as of July 2025 paid for Plus ($20/month) or Pro ($200/month) subscriptions, [3]according to The Information . The company reportedly projects that by 2030 about 8.5 percent of a weekly active user base of roughly 2.6 billion - around 220 million people - will pay for a Plus subscription. With so many non-paying customers, ads may be its only route to financial survival. (That or [4]a government bailout , which the company says won't happen.)
[5]
[6]
OpenAI has reportedly committed to [7]spend more than $1 trillion to fuel its quest for the vaguely defined idea of "artificial general intelligence." The AI giant lost north of [8]$11.5 billion in the third quarter alone, according to numbers Microsoft – a large OpenAI investor – disclosed in its financial filings.
Meanwhile, online advertising is a massive business, contributing more than $74 billion to Google's revenues and more than $50 billion to Meta's revenues in the third quarter of 2025 alone. And chatbots like ChatGPT are technically capable of capturing a massive amount of extremely personal data, given that people use them for everything from vibe coding to online companionship to informal therapy.
[9]
Simo acknowledged the need to preserve people's trust in ChatGPT – to the extent that people trust LLMs known for making things up and getting things wrong – and made a number of vague commitments to offer reassurance that their artificial intelligence agent wouldn't become a mercenary sales agent.
This included a reiteration of the claim that the company's pursuit of advertising aligns with its mission to ensure AGI – a moving target based on some determination of human intellectual equivalency – "benefits all of humanity."
[10]
OpenAI mockup of advertising in ChatGPT - Click to enlarge
Ads, Simo promised, will not influence the answers provided by ChatGPT, will be "optimized based on what's most helpful to you," as Simo put it, and will be distinct and clearly labelled.
ChatGPT users under 18, and queries about sensitive topics like health or politics, will be spared from ads, or so it's claimed.
"We keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers," Simo said, adding that users will have the ability to control how their data is used and to disable personalization.
[11]Experiment suggests AI chatbot would save insurance agents a whopping 3 minutes a day
[12]Lawmakers urge FTC to probe Trump Mobile over 'deceptive' marketing
[13]OpenAI invests in brain-interface biz co-founded by CEO Sam Altman
[14]Over half of AI projects are shelved due to complex infrastructure
"Selling data" can mean different things in different contexts, but may be disallowed if consent is lacking, depending on the relevant privacy rules. Under the California Consumer Privacy Act ( [15]CCPA ), it can mean "selling, renting, releasing, disclosing, disseminating, making available, transferring, or otherwise communicating orally, in writing, or by electronic or other means, a consumer's personal information by the business to another business or a third party for monetary or other valuable consideration."
Google and Meta have both insisted they [16]don't sell [17]personal data , but they certainly use it to target their own ads, and their decades of [18]privacy [19]missteps attest to the relevance of those claims.
[20]
Miranda Bogen, director of the Center for Democracy and Technology's AI Governance Lab, argues that even in the absence of data sharing, ad targeting can harm privacy. OpenAI's shift toward advertising presents risks for the company and its customers, she said.
"Even if AI platforms don't share data directly with advertisers, business models based on targeted advertising put really dangerous incentives in place when it comes to user privacy," said Bogen in an emailed statement. "This decision raises real questions about how business models will shape AI in the long run."
Bogen said that people use chatbots for a variety of reasons, including as companions and advisors. She argues that when tools try to exploit trust to sell goods, that may be problematic.
"AI companies should be extremely careful not to repeat the many mistakes that have been made — and harms that have resulted from — the adoption of personalized ads on social media and around the web," she said.
Meta has [21]integrated AI into its advertising business and [22]uses AI-generated images in ads . But efforts by chatbot maker Perplexity [23]to present ads in conjunction with its AI search service have fallen flat. Perplexity [24]paused accepting new ad clients in October last year, following the departure of its head of advertising.
Lower-tier ChatGPT users who haven't bought their way out of advertising may still have some say in whether they see ads. Several makers of ad blocking applications and extensions already [25]claim their tools can block ads that appear in ChatGPT. That may hold true when ChatGPT is accessed from the browser but is likely to be less feasible in OpenAI's native desktop apps. The possibilities will become clearer once OpenAI starts serving ads. ®
Get our [26]Tech Resources
[1] https://openai.com/index/our-approach-to-advertising-and-expanding-access/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://www.theinformation.com/articles/openai-projected-least-220-million-people-will-pay-chatgpt-2030
[4] https://www.theregister.com/2025/11/06/openai_cfo_walks_back_remarks/
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[7] https://blog.carnegieinvest.com/the-risks-facing-openai-and-its-1.4t-in-spending-commitments
[8] https://www.theregister.com/2025/10/29/microsoft_earnings_q1_26_openai_loss/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://regmedia.co.uk/2026/01/16/openai_ad_image.jpg
[11] https://www.theregister.com/2026/01/16/axlerod_ai_saves_insurance_agents_time/
[12] https://www.theregister.com/2026/01/16/ftc_trump_mobile/
[13] https://www.theregister.com/2026/01/16/openai_merge_labs_brain_interface_investment/
[14] https://www.theregister.com/2026/01/15/half_ai_projects_shelved/
[15] https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?division=3.&part=4.&lawCode=CIV&title=1.81.5
[16] https://www.facebook.com/privacy/policy/
[17] https://safety.google/intl/en_sg/safety/ads-data/
[18] https://en.wikipedia.org/w/index.php?title=Privacy_concerns_with_Google&oldid=1332407115
[19] https://en.wikipedia.org/w/index.php?title=Privacy_concerns_with_Facebook&oldid=1316949620
[20] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[21] https://www.facebook.com/business/ads/meta-advantage-plus
[22] https://www.meta.com/help/artificial-intelligence/355108217670024/
[23] https://www.perplexity.ai/hub/blog/why-we-re-experimenting-with-advertising
[24] https://www.adweek.com/media/perplexity-pauses-new-advertising-deals-to-reassess-ambitions/
[25] https://www.noadsgpt.online/blog/block-chatgpt-ads-2026
[26] https://whitepapers.theregister.com/
ChatGPT Go, the company's cheapest paid offering, has been an option in 171 countries since August and is now available stateside for $8 per month.
"In the coming weeks, we're also planning to start testing ads in the US for the free and Go tiers, so more people can benefit from our tools with fewer usage limits or without having to pay," [1]said Fidji Simo, OpenAI's CEO of applications. "Plus, Pro, Business, and Enterprise subscriptions will not include ads."
[2]
About 5 percent, or 35 million, of ChatGPT's weekly active user base as of July 2025 paid for Plus ($20/month) or Pro ($200/month) subscriptions, [3]according to The Information . The company reportedly projects that by 2030 about 8.5 percent of a weekly active user base of roughly 2.6 billion - around 220 million people - will pay for a Plus subscription. With so many non-paying customers, ads may be its only route to financial survival. (That or [4]a government bailout , which the company says won't happen.)
[5]
[6]
OpenAI has reportedly committed to [7]spend more than $1 trillion to fuel its quest for the vaguely defined idea of "artificial general intelligence." The AI giant lost north of [8]$11.5 billion in the third quarter alone, according to numbers Microsoft – a large OpenAI investor – disclosed in its financial filings.
Meanwhile, online advertising is a massive business, contributing more than $74 billion to Google's revenues and more than $50 billion to Meta's revenues in the third quarter of 2025 alone. And chatbots like ChatGPT are technically capable of capturing a massive amount of extremely personal data, given that people use them for everything from vibe coding to online companionship to informal therapy.
[9]
Simo acknowledged the need to preserve people's trust in ChatGPT – to the extent that people trust LLMs known for making things up and getting things wrong – and made a number of vague commitments to offer reassurance that their artificial intelligence agent wouldn't become a mercenary sales agent.
This included a reiteration of the claim that the company's pursuit of advertising aligns with its mission to ensure AGI – a moving target based on some determination of human intellectual equivalency – "benefits all of humanity."
[10]
OpenAI mockup of advertising in ChatGPT - Click to enlarge
Ads, Simo promised, will not influence the answers provided by ChatGPT, will be "optimized based on what's most helpful to you," as Simo put it, and will be distinct and clearly labelled.
ChatGPT users under 18, and queries about sensitive topics like health or politics, will be spared from ads, or so it's claimed.
"We keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers," Simo said, adding that users will have the ability to control how their data is used and to disable personalization.
[11]Experiment suggests AI chatbot would save insurance agents a whopping 3 minutes a day
[12]Lawmakers urge FTC to probe Trump Mobile over 'deceptive' marketing
[13]OpenAI invests in brain-interface biz co-founded by CEO Sam Altman
[14]Over half of AI projects are shelved due to complex infrastructure
"Selling data" can mean different things in different contexts, but may be disallowed if consent is lacking, depending on the relevant privacy rules. Under the California Consumer Privacy Act ( [15]CCPA ), it can mean "selling, renting, releasing, disclosing, disseminating, making available, transferring, or otherwise communicating orally, in writing, or by electronic or other means, a consumer's personal information by the business to another business or a third party for monetary or other valuable consideration."
Google and Meta have both insisted they [16]don't sell [17]personal data , but they certainly use it to target their own ads, and their decades of [18]privacy [19]missteps attest to the relevance of those claims.
[20]
Miranda Bogen, director of the Center for Democracy and Technology's AI Governance Lab, argues that even in the absence of data sharing, ad targeting can harm privacy. OpenAI's shift toward advertising presents risks for the company and its customers, she said.
"Even if AI platforms don't share data directly with advertisers, business models based on targeted advertising put really dangerous incentives in place when it comes to user privacy," said Bogen in an emailed statement. "This decision raises real questions about how business models will shape AI in the long run."
Bogen said that people use chatbots for a variety of reasons, including as companions and advisors. She argues that when tools try to exploit trust to sell goods, that may be problematic.
"AI companies should be extremely careful not to repeat the many mistakes that have been made — and harms that have resulted from — the adoption of personalized ads on social media and around the web," she said.
Meta has [21]integrated AI into its advertising business and [22]uses AI-generated images in ads . But efforts by chatbot maker Perplexity [23]to present ads in conjunction with its AI search service have fallen flat. Perplexity [24]paused accepting new ad clients in October last year, following the departure of its head of advertising.
Lower-tier ChatGPT users who haven't bought their way out of advertising may still have some say in whether they see ads. Several makers of ad blocking applications and extensions already [25]claim their tools can block ads that appear in ChatGPT. That may hold true when ChatGPT is accessed from the browser but is likely to be less feasible in OpenAI's native desktop apps. The possibilities will become clearer once OpenAI starts serving ads. ®
Get our [26]Tech Resources
[1] https://openai.com/index/our-approach-to-advertising-and-expanding-access/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://www.theinformation.com/articles/openai-projected-least-220-million-people-will-pay-chatgpt-2030
[4] https://www.theregister.com/2025/11/06/openai_cfo_walks_back_remarks/
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[7] https://blog.carnegieinvest.com/the-risks-facing-openai-and-its-1.4t-in-spending-commitments
[8] https://www.theregister.com/2025/10/29/microsoft_earnings_q1_26_openai_loss/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://regmedia.co.uk/2026/01/16/openai_ad_image.jpg
[11] https://www.theregister.com/2026/01/16/axlerod_ai_saves_insurance_agents_time/
[12] https://www.theregister.com/2026/01/16/ftc_trump_mobile/
[13] https://www.theregister.com/2026/01/16/openai_merge_labs_brain_interface_investment/
[14] https://www.theregister.com/2026/01/15/half_ai_projects_shelved/
[15] https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?division=3.&part=4.&lawCode=CIV&title=1.81.5
[16] https://www.facebook.com/privacy/policy/
[17] https://safety.google/intl/en_sg/safety/ads-data/
[18] https://en.wikipedia.org/w/index.php?title=Privacy_concerns_with_Google&oldid=1332407115
[19] https://en.wikipedia.org/w/index.php?title=Privacy_concerns_with_Facebook&oldid=1316949620
[20] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aWsXday3IhlD6cYrxJ68_gAAAtM&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[21] https://www.facebook.com/business/ads/meta-advantage-plus
[22] https://www.meta.com/help/artificial-intelligence/355108217670024/
[23] https://www.perplexity.ai/hub/blog/why-we-re-experimenting-with-advertising
[24] https://www.adweek.com/media/perplexity-pauses-new-advertising-deals-to-reassess-ambitions/
[25] https://www.noadsgpt.online/blog/block-chatgpt-ads-2026
[26] https://whitepapers.theregister.com/
Contrary to my previous conclusion…
Bebu sa Ware
you can actually enshittify a steaming pile of turds.
I, to my horror can now imagine: "would like fries with that" and this fresh hell "goes better with coke" … only conceivably when in powder form.
On the brighter side it might be the harbinger of decline and fall - the merest particle of hope at the bottom of this cesspool.
I gotta ask ...
jake
... who didn't see this coming?
Gotta squeeze as much as possible out of the suckers before the bubble bursts.
I might be accused of being a little jaundiced but.....
Anonymous Coward
So, in addition to being trained on a toxic torrent of turgid twaddle from t'internet in general, the product is now to be further enshittified by the addition of a slithering cesspool of sycophantic sponsored slime.
It's a bit like the jump from the original Clippy to the Clippy on steroids, with paedophilic tendencies, that is the AI offering and now discovering that clippy also has a severe Coke habit.
So we've now gone from GIGO to GIGOWAS (Garbage In, Garbage out, With Added Shit).
I doubt the universe is big enough to accommodate the barge pole I wouldn't touch this with.
Which ads?
- hopefully not the "magic heater" that keeps popping on YouTube even if you report it. It seems that Google would not use its own AI to block it. Or allows such ads intentionally.
Scam obviously outbid ads from real businesses with marginal value propositions. Could galloping inflation and lost productivity be due to major sectors of economy and talent devoted to scam, since the latter is much more profitable?