OpenAI's Ad Business Hits $1 Billion Annualized but Is Set to Miss Its Own $2.4 Billion 2026 Target

When OpenAI switched on advertising inside ChatGPT in February, it looked like an experiment bolted onto a subscription business. Ten months later it is a division with its own org chart, its own talent war and its own growth problem — just not the one competitors might hope. Ad revenue reached an annualized run rate of about $1 billion by the end of August, according to a detailed report carried by Sina Finance and other outlets citing people familiar with the figures. Against OpenAI's total annualized revenue of nearly $50 billion in late September, ads remain small. They are also growing faster than almost any ad channel the industry has seen.

The awkward part is the internal scoreboard. Early this year OpenAI told investors it expected $2.4 billion in total ad revenue for 2026. Despite the influx of advertisers, the company is now unlikely to hit that number, two people familiar with the matter said. Months of product launches for advertisers and successive price cuts brought volume through the door faster than dollars.

The volume, to be clear, is real. Research firm New Street Research counted roughly 60,000 advertisers on the platform in early October, triple the 20,000 of late August. Sensor Tower estimates U.S. ChatGPT ad spending will rise about 40% month over month in October. The business now operates in more than 60 countries. At New York Advertising Week last week, a joint session with Home Depot packed the room, with dozens of attendees lining up afterward to speak briefly with Dave Dugan, OpenAI's VP of global ad solutions. "We are building the fastest-growing ad business in the world," Dugan said, "hiring talent from across the industry rather than a single company."

The hiring has a distinct accent. About 140 employees work on ads, and nearly half joined between July and September, according to headcount tracker Crustdata. Dugan himself spent 12 years at Meta. Vijay Raj, who became CTO of applications a year ago, did a decade there; Mitch Piller (12 years at Google) and Andy Sinn (about 10 at Meta) joined in July to sell to brands and agencies in Australia and New Zealand; Mark Banbury, a veteran of more than ten years at X, took over EMEA ad sales at the end of September. Fidji Simo, who spent over a decade at Facebook before joining OpenAI, stepped back from full-time duties in early July, after which Dugan reorganized the team into vertical units for retail, travel and finance — mirroring the structure of the mature ad sales organizations at Meta and Google. In all, the ad team has been through seven reorganizations since February, according to two people familiar with the matter.

Budgets tell the more sobering story. Most advertisers are still running test allocations — typically 1% to 5% of annual ad spending — on ChatGPT, with only a handful scaling up meaningfully. Kate Scott-Dawkins, global head of research at GroupM, said contracting, onboarding and share-shifting all take time, and that Google will remain the dominant player in AI chat and AI search advertising for some time. She added, though, that she had never seen any ad channel grow as fast as OpenAI's.

The run-rate figures are company-reported and have not been independently audited, but the strategic reading does not depend on precision. OpenAI needs a second revenue pillar as subscription growth matures and its infrastructure bills compound, and advertising is the only mechanism that monetizes its roughly 1.2 billion weekly users at scale. Taking share from a market split among Google, Meta and Amazon was always going to be a multi-year fight. The first year's lesson is that demand arrived faster than budgets — a better problem than the reverse, but still a miss against a number the company itself chose.