As OpenAI pivots to advertising, it is becoming clear that our current AI revolution will rapidly expand the surveillance economy: the mass collection and monetization of our personal behavioral data.
Although many of us saw advertisements in LLMs to be an inevitability, OpenAI CEO Sam Altman had hoped to avoid this outcome. On more than one occasion, Altman went on the record to lambast ads. He said, “[Advertisements] fundamentally misalign a user’s incentives with the company providing the service,” and “I kind of think of ads as like a last resort for us as a business model.” And yet, here we are. Ads are coming to his LLM.
OpenAI’s financial quandary
Based on leaked shareholder disclosures, OpenAI allegedly had an operating loss of roughly $8 billion (excluding equity charges) in the first half of 2025. What’s more, the company says it will spend at least $600 billion in compute spend over the next five years, as it has agreements in place with Nvidia, AMD, Broadcom, and Oracle to buy data center capacity, chips, and other related infrastructure.
As of late February 2026, OpenAI had 900 million weekly active users; however, only 5.5% (50 million) pay for subscriptions. Venture capitalist and early Facebook investor Roger McNamee describes OpenAI’s pivot to advertising as a “five-alarm fire.”
According to the Financial Times, OpenAI intends to charge advertisers $60 per 1,000 impressions, in an effort to bring in billions in ad revenue in 2026. To be fair, other than ads, there are alternative revenue streams that OpenAI can pursue, including taking commissions on e-commerce transactions powered by ChatGPT; releasing wearables, and of course, raising funds via an IPO. OpenAI hopes to raise as much as $60 billion through a public offering, although it remains to be seen how accommodating the capital markets will be.
In a January New York Times essay, Sebastian Mallaby, a fellow at the Council on Foreign Relations, predicts that OpenAI will run out of money “over the next 18 months.” Mallaby suspects that it will be acquired by a well-heeled entity like Microsoft, which already owns roughly 27% of OpenAI via a cloud services partnership, or Amazon, which recently invested $50 billion in OpenAI.
Further complicating OpenAI’s financial situation, the company’s business model has virtually no moat.
OpenAI’s competitors
For its part, Google hasn’t integrated ads into Gemini yet, though ads are already popping up alongside its AI-generated search results (Google AI mode and the AI overview summaries for traditional Google searches). However, unlike OpenAI, Google can afford to hold off on advertising in Gemini because it already has plenty of revenue from its legacy businesses.
Another competitor, Perplexity, has experimented with ads, but it ultimately decided to pull the plug on them, declaring that ads were “misaligned with what the users want.” Perplexity very well may revisit ads down the line. In fact, I suspect they will.
And as many of us saw during its $8 million, 30-second Super Bowl spot, Anthropic has positioned itself as the ad-free alternative to ChatGPT. That said, even Anthropic CEO Dario Amodei is leaving the door open to ads. In his February 4th blog post, Amodei writes, “Should we need to revisit this approach, we’ll be transparent about our reasons for doing so.” This doesn’t exactly inspire confidence.
I believe that an advertising ecosystem with mass collection of user data is eventually coming to all of these LLMs—and a new wave of digital advertising is on the horizon.
The fourth wave of digital advertising
The first wave of digital advertising was search advertising and banner ads. With the second wave came social media advertising and mobile ads. And the third wave was comprised of large-scale data surveillance fueled by third-party cookies, cross-site tracking, and behavioral profiling—all in an effort to fuel advertising at scale. Within this third wave, it was the data brokers that really allowed behavioral targeting to scale.
Due to new laws (e.g., GDPR, CPRA), AI developments, declining third-party data quality, and sunsetting cookies, we’re now seeing the emergence of a fourth wave of digital advertising. This AI-fueled fourth wave is more reliant on first-party and zero-party data (data that is explicitly provided by the users).
This fourth wave is a shift to walled gardens and AI-native environments; through new chatbots and agentic shopping flows, customers will encounter ads based on their behavioral data. And without a doubt, this AI-fueled fourth wave will supercharge the surveillance economy.
Surveillance capitalism
The surveillance economy is closely related to Harvard Business School professor Shoshana Zuboff’s concept of surveillance capitalism. In her book, The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power, Zuboff describes surveillance capitalism as “a new economic order that claims human experience as free raw material for hidden commercial practices of extraction, prediction, and sales.”
Put differently, surveillance capitalism describes how entities acquire our behavioral data (frequently unbeknownst to us) and use it to generate revenue, often by selling it to data brokers and other third parties that use our data to send us targeted ads.
Big Tech’s widespread collection and commoditization of our personal data perfectly encapsulates surveillance capitalism. For years, our data has been tracked, aggregated, and sold in various markets; the largely unregulated data brokerage industry has ballooned to a $315 billion industry. And the current generative AI boom will further expand it.
Gen AI tools will fuel surveillance capitalism
According to a Fall 2025 report from the Harvard Kennedy School, “AI technologies—especially predictive models, automated decision systems, and generative tools feed on the data extraction pipelines that surveillance capitalism built.”
With each passing day, we see a further conflation of AI developments and the broader surveillance economy. As Harvard Kennedy fellow Burcu Kilic writes, “AI provides a new layer of legitimacy framed as progress, innovation, or national competitiveness, even as it deepens asymmetries of power and concentration of market and control. The surveillance capitalist order now markets itself as an AI revolution.”
Of course, hypothetically, this AI-fueled surveillance economy could always be reined in by federal data privacy legislation. But I’m not holding my breath.
