$60 per thousand impressions! OpenAI opens the curtain on "AI advertising" at a high price

OpenAI CEO Sam Altman, who once openly said that advertising made him “uneasy,” is now making major compromises in the face of business realities. To sustain the high current valuation of its AI platform and cover massive capital expenditures, OpenAI officially began ad tests on ChatGPT this week. This move not only marks a fundamental shift in the company’s business model, but also signals a coming shake-up in the global digital advertising market triggered by generative AI.

According to the Financial Times, OpenAI has set a high threshold for advertisers, with a starting price of $60 CPM (cost per thousand impressions) and a minimum spend requirement of $200,000. This aggressive pricing strategy is comparable to the level Netflix set when it launched its ad services at the end of 2022, showing that OpenAI aims to position its platform as a premium advertising channel and plans to generate billions in revenue through this business in 2024.

Currently, global top-tier advertising agency groups including WPP, Omnicom, and Dentsu have become early partners, involving clients from retail, software, travel and other industries. For advertisers, the commercialization of ChatGPT opens a whole new field. Marketers are eager to capture the consumer’s mind at the moment they seek answers through conversational AI. This transformation from traditional search engines to “answer engines” is reshaping the landscape of digital marketing.

However, this transition is not without risks. In its bid to monetize via ads and compete fiercely with rivals like Google and Anthropic, OpenAI must find a delicate balance between revenue growth and user trust; a misstep could undermine its core product’s value proposition.

Pricing Strategies and Business Compromises

For OpenAI, introducing ads is a critical step in its transformation from a research lab into a commercial giant. Just 18 months ago, Altman made it clear in a Harvard speech that he “hates advertising,” believing it fundamentally distorts the incentive mechanism between users and service providers. However, in the face of rising compute and infrastructure costs in this AI arms race, finding new sources of revenue has become imperative.

According to insiders, OpenAI is currently testing only for free users in the U.S., with highly “selective” pricing. Stagwell Europe Chief Growth Officer James Denton-Clark noted that the minimum investment threshold for early pilots is set at six figures ($200,000), so early demand is mainly from large, mature advertisers. These ads are shown separately from ChatGPT-generated answers, similar to traditional billboards, and designed to build brand awareness.

Jessica Tamsedge, CEO of Dentsu Creative UK&I, said this is an “obvious” opportunity for advertisers. She especially mentioned Walmart’s stock surge after announcing its partnership with OpenAI as proof of market recognition of AI’s commercial potential. Forrester Chief Analyst Nikhil Lai said, OpenAI is taking a deliberate gamble, betting that ChatGPT’s high engagement will justify the high upfront costs for advertisers.

Challenging the Traditional Search Giants

OpenAI’s entry directly threatens the digital ad market long dominated by tech giants like Google and Amazon; the market size has surpassed $1 trillion. Kate Scott-Dawkins, WPP Media’s Global Head of Business Intelligence, pointed out that search engines which defined the last generation of ad models are being replaced by conversational AI. Because AI platforms can offer highly precise suggestions based on users’ past behavior (e.g., translation history or shopping preferences), traditional search and e-commerce platforms may lose ad revenue.

Despite Google currently leading the industry with over $80 billion in quarterly ad revenue and introducing AI summary ads into traditional search interfaces, it has not placed ads in its Gemini chatbot. Analysts believe Google can afford to wait because it has a vast existing ad business as a cushion, giving it an edge in appealing to users who dislike ads.

By contrast, OpenAI lacks this historical buffer. Consulting firm Madison and Wall’s analysis shows that almost all internal cash flow at tech companies is being reinvested in AI infrastructure, leaving little room for buybacks or dividends. Madison and Wall Managing Director Luke Stillman described OpenAI's current strategy as “throwing spaghetti at the wall,” trying every possible way to create more revenue.

Rival Counterattacks and Trust Concerns

While OpenAI is actively embracing ads, its main competitor Anthropic has taken a starkly different stance. About 80% of Anthropic’s expected revenue comes from enterprise clients, versus only 40%–50% for OpenAI. After OpenAI announced its ad plan, Anthropic ran an ad during the Super Bowl stating: "Advertising is moving into AI, but it won’t come to Claude" (referring to its chatbot product).

This differentiated strategy highlights industry concerns around user privacy and trust. Cornell Professor David Rand warned that if users worry their personal information is used for ad targeting, they may become more cautious when conversing with ChatGPT, directly reducing the utility of AI. To alleviate this concern, OpenAI promises not to use private chat records for ad targeting, but only to match relevance based on the user’s current prompt, and to allow free users to opt-out within certain limits.

Despite the controversy, advertising executives generally believe that as consumer habits change, chatbots will be the next frontier in marketing. Although WPP estimates OpenAI’s first-year ad revenue may reach $500 million–$800 million—still far below its multi-billion dollar internal goals—this is seen as a strong start. As former WPP CEO Mark Read said, logic suggests a hybrid model of subscriptions and ads will work best for AI apps, just as Netflix and Amazon have proven.

Risk warning and disclaimerThe market involves risks, and investment should be approached cautiously. This article does not constitute personal investment advice, nor does it take into account the special investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article fit their own circumstances. Investments made based on this information are at your own risk.