OpenAI announced on Friday the upcoming arrival of advertising in ChatGPT, starting with a test phase in the United States for discounted subscribers and free users of the very popular artificial intelligence (AI) assistant, under pressure to make its massive expenditure profitable.
“In the coming weeks, we are preparing to start testing advertising in the United States for free users and the Go offer”, first-price subscription, announced OpenAI, confirming a highly anticipated shift that has been commented on for weeks in Silicon Valley.
“Plus, Pro and Enterprise subscriptions will not include advertising,” the group specifies in a long blog post detailing how it intends to introduce this source of revenue while “maintaining the trust” of users in the responses of ChatGPT, the most used AI conversational agent in the world.
With only a fraction of its billion users under subscription, OpenAI is under pressure to generate new revenue.
If its valuation has climbed to 500 billion dollars in private funds since 2022, and an IPO worth 1,000 billion is mentioned, the group is burning through its resources at an alarming speed. At issue: the colossal cost of the computing power required to run the AI.
By taking the plunge, OpenAI is aligning its model with that of the giants Google and Meta, whose power relies primarily on advertising backed by their free services.
Unlike OpenAI, these giants draw on these advertising revenues, generating tens of billions of dollars in annual profits, to carry out their ambitions in AI. Amazon is following a similar trajectory, with a growing advertising business on its sales and video services.
– “monetize attention”? –
“Advertising is not a distraction in the generative AI race; it’s OpenAI’s way to stay there,” said Jeremy Goldman, an analyst at Emarketer.
“OpenAI admits something simple and far-reaching: competition is no longer just about the quality of models, but about the ability to monetize attention without sacrificing trust,” he added.
“Our priority is user trust and experience before revenue,” promises the signator of the post, French leader Fidji Simo, right-hand man of OpenAI boss Sam Altman.
To convince, the manager details the principles supposed to govern this introduction: “Ads do not influence responses”, conversations “remain private”, data “never sold to advertisers”, etc.
The tested advertisements will be displayed at the bottom of ChatGPT responses, “clearly identified and distinct” from them.
“We do not optimize the time spent on ChatGPT,” continues Fidji Simo, while a number of tech companies, such as Meta (Facebook, Instagram), TikTok and Snapchat, are accused of designing their algorithms to keep users as long as possible on these platforms, whose profitability relies on advertising.
The commitment to preserving the well-being of users is also a sensitive subject for OpenAI, accused of having given priority to emotional involvement and freedom of use to the detriment of users’ mental health. Several legal proceedings targeting OpenAI and other tech giants are underway in the United States, particularly in connection with cases of suicide.
Silicon Valley is also buzzing with speculation about the introduction of advertising by ChatGPT’s competitors, in particular for the increasingly popular Google Gemini.
There are “no plans for advertising in the Gemini app,” Dan Taylor, Google’s vice president of global advertising, responded in an interview last week with Business Insider.
