[CEONEWS=Joon-cheol Bae, Repoter] “The combination of AI and advertising feels uniquely unsettling. Advertising is our last resort.”
That was what Sam Altman, CEO of OpenAI, said at a Harvard University event in May 2024.
Twenty months later, on January 16, 2026, that “last resort” became reality. OpenAI officially announced that it would introduce advertising to the free version of ChatGPT and to its $8-per-month low-cost tier, ChatGPT Go. On the day of the announcement, Altman wrote on X (formerly Twitter):
“Many people clearly want to use AI but don’t want to pay for it. We hope this business model works.”
It was the moment when a leader who had long spoken in ideals chose to compromise with reality.
■ An $8 Billion-a-Year “Cash Furnace”… Advertising Was Inevitable
The decision to introduce ads was hardly optional. The Financial Times famously described OpenAI as an “era-defining money furnace.” In 2025 alone, the company reportedly burned through roughly $8 billion (about KRW 11 trillion) in cash.
ChatGPT boasts around 800 million weekly active users, yet only about 5% are paid subscribers. The vast majority of users are free riders, while the server and electricity costs they generate are astronomical. As AI models grow more advanced, inference costs rise exponentially. Pressure from major investors, including Microsoft, has intensified. At this stage, visions of “changing the future” are no longer enough to justify trillion-won losses.
■ Ads “Separated From Answers, No Sale of User Data”
OpenAI insists its ad model will differ fundamentally from conventional internet advertising. Rather than disruptive banner ads, sponsored products or services will appear beneath responses, aligned with the context of the user’s query.
Fidji Simo, CEO of OpenAI Applications, outlined the company’s advertising principles in a blog post:
Ads will not influence ChatGPT’s answers.
User conversation data will not be sold to advertisers.
Ads will be clearly labeled and separated from organic responses.
Users will be able to see why a specific ad appeared and hide it if desired. Accounts confirmed or inferred to belong to users under 18 will not see ads, nor will ads appear near sensitive topics such as health, mental health, or politics.
Higher-tier subscriptions—ChatGPT Plus ($20/month), Pro, Business, and Enterprise—will remain completely ad-free.
■ From Idealist to Pragmatic Operator
Altman’s shift in tone is striking. In 2024, he went as far as saying he “aesthetically disliked” advertising. But by the second half of 2025, his stance began to change. In an interview with Stratechery, he remarked, “I like Instagram ads,” adding that they help him discover products he wouldn’t have found otherwise. He went on to say that “great ad products that generate net value for users” are possible.
Industry observers interpret this as “a Silicon Valley idealist embracing Wall Street logic.” Innovation cannot continue if the company itself cannot survive. With competitors like Google’s Gemini and Anthropic’s Claude aggressively chasing market share, maintaining a technological lead without a reliable revenue stream is nearly impossible. In this view, advertising is not a betrayal of philosophy but fuel to sustain long-term innovation.
■ The AI Industry Enters the “Platform Business” Era
Experts see this move as a signal that the AI industry is entering an early stage of maturity. Just as Google made search free and monetized attention through ads, and Meta built open social networks supported by advertising, OpenAI is now following the textbook platform-business path.
Jeremy Goldman, an analyst at eMarketer, commented:
“Advertising isn’t a sign of exiting the generative AI race—it’s how you stay in it. If ChatGPT turns on ads, OpenAI is acknowledging a simple but crucial reality.”
He added, “The competition is no longer just about model quality. It’s about monetizing attention without destroying trust.”
The AI service market is rapidly splitting into a dual structure:
Free or low-cost, ad-supported services for the mass public, and
Premium, ad-free subscription services for enterprises and professionals.
This shift reflects a deeper truth: AI is no longer a novelty. It has become everyday infrastructure.
■ The Remaining Challenge Is Trust… If Ads Distort Answers, It’s Over
The greatest risk is erosion of trust. If AI responses are perceived as biased in favor of advertisers, users may begin to question whether they are reading facts or paid persuasion.
Inside OpenAI, the move reportedly sparked internal resistance. Some employees expressed concern that advertising conflicts with the company’s mission to achieve artificial general intelligence (AGI).
Altman has firmly stated that ads will never influence ChatGPT’s answers. Whether that promise holds will be judged over time. If advertising degrades the user experience, cynicism—“money always wins in the end”—will spread, and users may migrate to alternatives.
Will this “last resort” become the worst possible handshake—or a masterstroke?
CEOs around the world are watching Sam Altman’s next move closely.
Copyright ⓒ CEONEWS 무단 전재 및 재배포 금지
본 콘텐츠는 뉴스픽 파트너스에서 공유된 콘텐츠입니다.
