OpenAI is making the case that advertising can fund ChatGPT for people who do not pay, while proposing to match ads to what users are asking. That puts the company’s promises about privacy and trust directly alongside a fast-growing advertising business.
OpenAI Watch analysis
What happened
At Advertising Week in New York, OpenAI vice-president of monetisation partnerships Samantha Jacobson said the company pursued ads so “all of humanity could benefit from AGI”. She argued that ads can support access for people unable to pay for ChatGPT, and suggested that what a person is trying to do matters more than a traditional profile built from personal attributes.
Digiday reports that OpenAI ads are matched to what someone asks and, unless personalisation is switched off, to signals from their wider use of the product. The company’s published ad principles, as described by Digiday, say conversations will remain private from advertisers and user data will not be sold to them. Jacobson said trust would matter even if protecting it came at the expense of revenue or partnership opportunities.
The business is moving quickly: OpenAI began selling ChatGPT ads in February and opened a self-serve Ads Manager in the US in May. Digiday says OpenAI announced a $1 billion annualised ad run rate at the end of August, with tens of thousands of advertisers and availability in more than 40 countries. The publication also reports that OpenAI told investors it expects $2.5 billion in ad revenue this year and is targeting $100 billion by 2030.
Why it matters
The question is not simply whether ads appear in a chatbot. It is what gets used to make them relevant, and whether people see a useful answer or an advert shaped by a private conversation. OpenAI’s distinction between identity and intent is commercially neat; users may still regard the contents of a chat as personal, whoever is doing the targeting.
The scale of the revenue ambitions makes the trust pledge consequential. A promise to keep advertisers away from conversations is meaningful, but readers will want to know how ad personalisation works in practice, what is switched on by default and how easily users can change it.
Our read
OpenAI has a credible reason to offer an ad-funded route for people who will not subscribe. But “we target the moment, not the person” is not a magic spell that makes conversational data feel impersonal. The test will be the controls users actually get, and whether the ads remain clearly separate from the answers they came for.
What to watch
- What information is used to match ads, and what personalisation settings are on by default.
- Whether OpenAI provides clear controls to review or switch off ad personalisation.
- Whether the company’s privacy promises hold as its advertising targets grow.
Discussion spark: If ads are matched to what you ask a chatbot, is that useful context or a step too far into private conversation?
Sources and evidence
- OpenAI would like to remind marketers that its ads are for the good of humanity – Digiday (8 October 2026, 04:07 UTC)
OpenAI Watch is independently operated by WittyWires. It is not affiliated with, endorsed by, or operated by OpenAI.