The New York Times sued Perplexity for copyright infringement—months after signing an AI licensing deal with Amazon. Perplexity built revenue-sharing programs for publishers. The Times declined to join any of them. Now lawyers are involved.
OpenAI has 800 million users. Only 5% pay. The company is bleeding billions quarterly while chasing a subscriber base ten times larger than Netflix just to break even. Now ads look inevitable—and user trust hangs in the balance.
Hugging Face's Skills tool lets Claude fine-tune competing models for thirty cents. A 7B parameter cap and subscription fees complicate the democratization pitch. The deeper issue: access to a button isn't access to understanding.
OpenAI has 800 million users. Only 5% pay. The company is bleeding billions quarterly while chasing a subscriber base ten times larger than Netflix just to break even. Now ads look inevitable—and user trust hangs in the balance.
They say desperation breeds innovation. For OpenAI, it breeds something closer to pragmatism dressed in Silicon Valley optimism. The company that once promised to democratize artificial intelligence is now—quietly, then less quietly—discussing ads. Not just any ads, mind you. We're talking about contextual, memory-based, personalized ads woven into the fabric of ChatGPT itself, rolling out sometime in 2026.
This isn't a pivot born from abundance. This is a company with 800 million users, only 5% of whom pay, now searching for a revenue model that actually works. And ads, it seems, are looking increasingly inevitable.
But what does this mean for user trust and investors betting on OpenAI's $500 billion valuation?
Here's where the plot thickens. The numbers suggest something uncomfortable: the LLM industry would need about 3 billion paying subscribers just to break even on infrastructure costs alone. To contextualize that absurdity—Netflix, after nearly two decades of dominance, has just 300 million paid subscribers. The industry was chasing a subscriber base 10 times larger than one of the world's most successful streaming platforms to merely survive, let alone profit.
OpenAI's situation perfectly illustrates this mathematical nightmare. The company generates roughly $13 billion in annual revenue but has lost more than $12 billion in a single quarter. Meanwhile, competitors like Anthropic are projecting profitability by 2027, having focused on enterprise customers with locked-in contracts rather than chasing the free user mirage. But their audacious leap from $4.7B in 2025 revenue to $70B by 2028 raises eyebrows about whether enterprise demand can truly scale that fast amid intensifying competition and compute crunches.
Ads, then, aren't a visionary move. They're a lifeline. And I wouldn't be surprised if Anthropic follows OpenAI down this path sooner than expected.
Look at how Netflix and YouTube evolved, and you'll see OpenAI's future written in the margins.
Some users already assume ChatGPT's answers are ranked by sponsorship. The trust deficit is real, and it's already baked into user perceptions.
Netflix's ad-supported tier, initially viewed skeptically, has become a powerhouse. Today, 42% of Netflix subscribers opt for the ad-supported plan—up from just 14% two years ago. These aren't fringe users either; they're often older, cost-conscious, and perfectly content to tolerate ads in exchange for affordability. Netflix is making this work because it created a tiered system where viewers choose their compromise. The ad-supported tier isn't cannibalistic; it's additive.
YouTube, meanwhile, has mastered the art of making you forget which side of the ad-supported fence you're on. The platform generates roughly $36.1 billion annually from advertising alone, with creators earning a cut while YouTube and Google keep substantial margins. YouTube Premium subscribers—who pay to eliminate ads—represent only a fraction of total viewership, yet the ad-supported model generates sufficient revenue to keep the entire ecosystem humming.
Spotify, operating on a similar freemium principle, pulls roughly 85% of its revenue from premium subscriptions, but its ad-supported free tier serves a critical function: it's the funnel through which premium conversions happen. The free tier with ads isn't a loss leader; it's a customer acquisition engine.
Here's where OpenAI's advertising dilemma becomes particularly thorny. Unlike Netflix or YouTube—platforms where users expect commercial interest—ChatGPT has cultivated an image of neutral counsel. It's the tool you turn to for unbiased information, objective analysis, and recommendations unmoored from profit motive.
CEO Sam Altman understands this danger. He's acknowledged that if ChatGPT accepted payment to rank a worse hotel above a better one, it could be "catastrophic for your relationship with ChatGPT." This isn't hyperbole. The trust deficit is real, and it's already baked into user perceptions. OpenAI focus groups have reportedly found that some users already assume ChatGPT's answers are ranked by sponsorship. The company hasn't even launched ads yet, and credibility erosion is already underway.
Yet Altman also claims OpenAI views ads as "something we may try at some point," and has framed advertising as not their "biggest revenue opportunity." This linguistic dance—simultaneous denial and inevitability—suggests internal conflict. The company needs revenue. It knows ads are coming. It's just hoping the messaging can soften the landing.
OpenAI sees ads as a core pillar, not a side hustle.
The financial projections are telling. OpenAI forecasts that advertising could generate $1 billion in revenue starting in 2026, ramping to nearly $25 billion by 2029. For context, that's only $4 billion less than what the company projects for enterprise AI agents—the crown jewel of its ambitions.
This suggests OpenAI sees ads as a core pillar, not a side hustle.
The question isn't whether ads are coming. They are. The question is what happens to user trust when the company discovers that monetizing free users through advertising creates a fundamental conflict with the promise of impartial intelligence.
Netflix and YouTube succeeded with ads because users understood the basic economic bargain: ads subsidize access. But AI assistants operate in murkier territory. When you ask ChatGPT for restaurant recommendations, you're implicitly asking for honesty. Inject ads into that equation, and you've transformed the tool from advisor to sales channel.
Spotify manages this by separating ad experience from premium experience. YouTube does similarly. But OpenAI's leaked code hints at something more integrated: ads embedded within the conversation itself, potentially shaped by user conversation history and preferences. This isn't a banner at the bottom of the page. This is advertising inside the advice.
That distinction matters enormously.
That belief becomes harder to maintain the moment users begin suspecting that the platform's core function—giving you information—has been colonized by commercial interest.
If you can't get 3 billion paying customers, monetize the 800 million free users you already have.
What's emerging, then, is a three-tier model: free with ads, ChatGPT Plus ($20/month ad-free), and ChatGPT Pro ($200/month for power users). Each tier would theoretically capture different user segments, much like Netflix's structure. The math suggests this could work—barely.
But here's the catch: it requires sustaining the belief that ChatGPT Plus is genuinely ad-free, genuinely better, and genuinely worth the premium. That belief becomes harder to maintain the moment users begin suspecting that the platform's core function—giving you information—has been colonized by commercial interest.
The LLM industry faced an impossible equation: 3 billion subscribers needed just to break even. OpenAI's solution is elegant in its desperation: if you can't get 3 billion paying customers, monetize the 800 million free users you already have. Ads are the pressure valve for an unsustainable growth model.
Ads could deliver $25 billion by 2029, yet they demand flawless execution to justify multiples exceeding 35x forward earnings.
OpenAI's pivot to ads signals a pragmatic revenue chase amid escalating losses, but it carries valuation risks for investors. The company hit a $500 billion valuation in October 2025 through employee share sales, yet projects $74 billion in operating losses by 2028—75% of revenue—driven by $1.4 trillion in committed compute spending over eight years. Ads could deliver $25 billion by 2029, yet they demand flawless execution to justify multiples exceeding 35x forward earnings, where even $425 billion in industry revenue barely supports current hype.
Anthropic offers a sharper contrast, targeting profitability by 2028 via 80% enterprise revenue, burning just 70% of its $4.2 billion 2025 sales versus OpenAI's equivalent rate on $13 billion. Investors betting on OpenAI face dilution risks from secondary sales and cloud commitments, while Anthropic's $183 billion valuation reflects steadier $9 billion run-rate growth through corporate predictability. A stumble here could trigger sector-wide multiple compression, hitting Nvidia and cloud giants as AI cash burn resets return expectations.
For the broader AI industry, ads validate hybrid models but expose overreliance on consumer scale. Success might stabilize OpenAI's lead, funneling capital to efficient players; failure risks a 33% P/E repricing across high-flyers, favoring enterprise-focused survivors over subscriber mirages. The real wager: can ads preserve the neutrality premium that inflated valuations, or will trust erosion force a painful reset?
For more insights about what AI can or cannot do, check out my book “Artificial Stupelligence: The Hilarious Truth About AI”.
The New York Times sued Perplexity for copyright infringement—months after signing an AI licensing deal with Amazon. Perplexity built revenue-sharing programs for publishers. The Times declined to join any of them. Now lawyers are involved.
Chinese hackers operated inside U.S. VMware servers for 17 months undetected. The malware repairs itself when deleted. It hides where most security teams don't look. CISA's December 4 advisory exposes an architectural blind spot in enterprise defense.
Werner Vogels ends his 14-year keynote streak by handing out printed newspapers and warning developers about "verification debt." His parting message: AI generates code faster than humans can understand it. The work is yours, not the tools.
Meta is cutting its metaverse budget by 30%. Wall Street cheered. But the money isn't disappearing—it's moving to AI glasses and wearables. Is this fiscal discipline or just the same spending habit with a new label?
Get the 5-minute Silicon Valley AI briefing, every weekday morning — free.
