A recent OpenAI job listing hints that the AI giant is laying the groundwork for an ad network.
So far, OpenAI’s nascent ad business has been a platform for advertisers looking to reach ChatGPT users. Now, there’s evidence that the AI giant might want to jump into a new side of the market — helping publishers and other companies sell their own ad space to marketers, CMO Insider can exclusively report.
The OpenAI job ad, posted last week, called for a “support delivery lead” in its ads division. The job posting, which OpenAI modified after CMO Insider asked about it, referenced building a publisher-facing ad business, the first time the company has publicly mentioned ambitions in the area.
Such a move could boost ad revenue, offer marketers placements beyond the ChatGPT app, and potentially soothe tensions with some publishers.
“This is a huge sign that OpenAI is building toward an advertising business far beyond just ChatGPT,” Debra Aho Williamson, founder of the AI-focused research and advisory firm Sonata Insights, told me.
The job posting originally said that the role’s publisher-side responsibilities would include “inventory setup, ad serving integrations, reporting, and yield-impacting issues.”
This person would “design and continuously improve support coverage for both ad buyers and publishers, ensuring the team can diagnose buyer-side delivery issues and publisher-side monetization and integration issues with equal rigor,” the job posting said.
OpenAI updated the job ad after CMO Insider contacted the company about it, removing references to publishers. The company deleted or edited five sentences in the posting. OpenAI later said the role’s responsibilities cover agencies and brands.
Intriguing!
So, what could an OpenAI ad network look like?
The job ad didn’t fully sketch out what OpenAI is interested in building. The most obvious move for OpenAI would be to enter the world of “demand extension” or “audience extension.” OpenAI could take growing advertiser demand and extend those campaigns to placements on a network of external websites or apps. Publishers would be able to monetize their own traffic using the commercial signals gathered by ChatGPT. Think of it as an AI-native equivalent to the Facebook Audience Network or Google Display Network.
OpenAI told CMO Insider it isn’t entering demand extension.
Strategically, creating an ad network would make sense for OpenAI, though it would represent an unusually fast jump into supply-side adtech.
One of the biggest advertiser gripes with OpenAI’s ad product so far is that advertiser demand has outpaced available supply. Advertisers previously complained that OpenAI had been slow to actually spend their budgets, despite asking for hefty upfront commitments to participate in early ad trials.
OpenAI’s head of global ad solutions, Dave Dugan, told me in a June interview that ad delivery and volume had improved as its confidence in the ChatGPT Ads user experience had grown. Still, OpenAI only serves ads to users on ChatGPT’s free or lower-priced tiers in select regions, meaning its native ad real estate remains fundamentally constrained by the chat interface.
A publisher ad network could help with the inventory bottleneck.
Building bridges with publishers
Launching a sell-side product would also mark a dramatic shift in OpenAI’s relationship with media companies.
To date, publishers do not receive a cut of ChatGPT’s ad revenue. Instead, some publishers make money from OpenAI through direct licensing deals for training data and content retrieval. (Business Insider parent company, Axel Springer, has one of these deals.)
The broader relationship remains tense: publishers including The New York Times, Ziff Davis, and a coalition of local news outlets are actively suing OpenAI, alleging copyright infringement. OpenAI has argued that its training of AI models is protected by the fair use doctrine.
An ad network would introduce a new way for OpenAI to compensate publishers, giving them a cut of ad revenue if a user views or clicks an ad served on their site or app.
As OpenAI prepares for an eventual public offering, it faces pressure to scale high-margin revenue streams, such as ads. The company must justify its eye-popping private valuation and offset hundreds of billions in long-term compute and infrastructure commitments.
In April, Axios reported that OpenAI was forecasting about $2.5 billion in ad revenue in 2026 and around $100 billion in 2030.
Nate Elliott, an analyst at BI’s sister company EMARKETER, said that for OpenAI to hit those numbers, it would need to dramatically expand its user base, charge top-tier ad rates, and potentially risk ruining the user experience with a heavy ad load.
Or, it could build an ad network.
Read the full article here
