OpenAI's Growing Gamble: Why Industry Veterans Question Its New Ad Strategy

When Sam Altman stood before Harvard University two years ago, he was categorical: advertising would be the company’s “last resort” as a revenue model. He warned that injecting ads into ChatGPT responses would erode user trust in OpenAI’s flagship product. Fast forward to today, and that conviction has crumbled under mounting financial pressure. The company just launched advertisements in ChatGPT—a direct reversal that highlights the deepest crisis facing the world’s most valuable AI startup: it’s running out of money faster than it can generate it.

The math is unforgiving. OpenAI generated roughly $13 billion in revenue in 2025, but the company plans to invest another $100 billion over the next four years—the bulk going toward the computing infrastructure necessary to train and deploy its AI models. The gap between income and outflow is widening dangerously. With public markets still skeptical and willing investors drying up globally, OpenAI must find new revenue streams urgently. But as it sprints toward profitability, it’s stretching itself across multiple unfamiliar battlegrounds simultaneously: advertising, enterprise software, scientific partnerships. The question haunting the industry is whether the company can execute across all three fronts without faltering.

The Advertising Experiment Nobody Wanted

Entering the ad business puts OpenAI in uncharted territory. The company has no track record in advertising—no sales infrastructure, no client relationships, no understanding of how ad buyers behave. Meanwhile, it’s competing against entrenched giants like Google, which has dominated enterprise advertising for decades.

Brian O’Kelley, CEO of Scope3 and a veteran of internet advertising with two decades of experience in the field, captured the tension perfectly when speaking to analysts: “OpenAI is trying to win over consumers, catch up with Anthropic’s programming tools, build data centers, and keep fundraising all at once. There’s just too much it’s chasing. Can it really do advertising well? Can it really do everything it wants to do well?” His skepticism reflects a broader industry concern—that OpenAI’s ambitions may exceed its capacity to execute.

The company has begun assembling an ad sales team, but progress remains sluggish. In May 2025, OpenAI hired Fidji Simo, a veteran of Facebook and former CEO of Instacart (where she oversaw a successful shift toward an advertising-focused business model), to lead the applications division. Since then, the company has poached hundreds of employees from Meta and X, many with ad product experience. Yet Mark Zagorski, CEO of DoubleVerify—a firm that works closely with major ad platforms—noted that OpenAI lacks the foundational infrastructure: “They need to build the technical systems required to operate an ad business.”

The Netflix comparison offers some perspective: it took the streaming giant two years to build a viable advertising operation, and even then, Netflix outsourced most of the heavy lifting to more experienced partners.

The Enterprise Market: Crowded and Costly

OpenAI’s second revenue pillar is enterprise software. The company wants to boost enterprise revenue from 40% of total income to 50% by year’s end. It’s offering tools like Codex (for developers) and ChatGPT Enterprise, with some enterprise customers paying up to $200 monthly.

Yet this market is increasingly hostile. Google and Microsoft bring decades of enterprise relationships and embedded ecosystems. More troubling is Anthropic, OpenAI’s closest rival, which has focused heavily on enterprise tools and is gaining traction with its Claude programming assistant, ClaudeCode. As if to underscore the competition, Anthropic recently aired a Super Bowl advertisement mocking OpenAI’s shift into ads—“The age of AI ads has arrived—but Claude has no ads.” Altman fired back on social media, defending the move as a way to bring AI to “billions who can’t afford a subscription.” But the insult landed: it exposed a strategic vulnerability. Ordinary enterprises, according to UBS analyst Karl Keirstead, may balk at the high fees OpenAI is asking for office software. OpenAI must compete on both price and capability against formidable opponents.

The Value-Sharing Gamble and Its Backlash

In late 2025, at the World Economic Forum in Davos, OpenAI CFO Sarah Friar introduced yet another revenue concept: “value sharing.” If OpenAI’s technology helps generate breakthrough discoveries—particularly in pharmaceuticals—the company might take a cut of resulting profits. The idea triggered immediate concern. When OpenAI subsequently launched Prism, a product aimed at scientists, many researchers cited Friar’s remarks and worried that OpenAI intended to claim ownership stakes in their discoveries.

The backlash forced OpenAI into damage control. Kevin Weil, the newly appointed Chief Science Officer, clarified on social media that the company would not extract profits from individual researchers. However, Weil left the door open for partnerships with large pharmaceutical firms where profits could be shared. Altman later echoed this stance, suggesting OpenAI might “explore partnership models where we bear costs and share in proceeds.” The episode illustrated how quickly unconventional business ideas can alienate customers—a risk OpenAI cannot afford as it attempts to diversify revenue.

The Core Challenge: Can OpenAI Do It All?

What emerges from this picture is a company in strategic overextension. OpenAI is simultaneously building a new advertising operation from scratch, fighting entrenched competitors in enterprise software, experimenting with equity-like arrangements in scientific research, building data centers, and trying to raise another $100 billion. Each initiative carries execution risk. Each demands specialized talent and operational expertise. And each competes for attention and resources.

Brian O’Kelley’s two-decade perspective on these dynamics carries weight. The reality he implicitly raises is whether any organization—regardless of funding or talent—can genuinely master multiple complex businesses at once. OpenAI’s pivot from Sam Altman’s stated principles suggests the company is not simply executing a coherent strategy but reacting to a widening crisis. That reaction may ultimately determine whether OpenAI remains the leader in AI or becomes a cautionary tale about scaling too fast without a sustainable business model.