When Apple unveiled its partnership with OpenAI to bring ChatGPT into Siri, it looked like a triumph for consumers—smarter voice assistants, better answers, fewer frustrations. But beneath the polished keynote and the promise of helpfulness lies a business model that should give every consumer and regulator pause: the companies building your AI assistant are, at their core, advertising companies. And the ad placements they’re engineering may be the most intimate and persuasive in the history of commerce.
A provocative analysis published by Juno Labs lays out the argument in stark terms. The thesis is straightforward: every major company racing to put an AI assistant in your pocket, your car, or your kitchen already derives the majority of its revenue from advertising, or is rapidly building the infrastructure to do so. Google, Meta, Apple, Amazon, Microsoft—these aren’t neutral technology providers. They are, functionally, advertising platforms with engineering departments.
The Revenue Tells the Story
Consider the numbers. Alphabet, Google’s parent company, generated over $307 billion in revenue in 2024, with approximately 77% coming from advertising. Meta’s advertising share is even more dominant, hovering near 97% of total revenue. Even Amazon, ostensibly an e-commerce company, has built an advertising business worth more than $46 billion annually—making it the third-largest digital ad platform in the United States. Microsoft, through LinkedIn and Bing, has been steadily growing its own ad revenue, which now exceeds $18 billion per year.
As Juno Labs argues, the financial gravity of these companies pulls everything—including AI assistants—toward advertising. When Google integrates Gemini into search, or when Meta embeds Meta AI into WhatsApp and Instagram, the underlying incentive structure hasn’t changed. These assistants are not being built out of altruism. They are being built because conversational AI represents the next great surface for monetization.
From Search Bars to Conversations: A New Advertising Frontier
The shift from traditional search advertising to conversational AI advertising represents a profound change in how commercial influence operates. In the old model, a user typed a query into Google, received a list of ten blue links, and some of those links were paid placements clearly marked as ads. The user retained agency—they could see the ads, skip them, and click on organic results instead.
In the conversational AI model, that transparency evaporates. When you ask an AI assistant to recommend a restaurant, a laptop, or a financial advisor, the response comes as a single, authoritative-sounding answer. There is no list of alternatives. There is no “Ad” label. There is just the assistant’s recommendation, delivered in a friendly, human-like tone. The Juno Labs analysis highlights this as the core danger: “When your AI assistant recommends a product, you won’t know if it’s because it’s the best option for you—or because someone paid for it to say that.”
The Trust Problem No One Is Talking About
This matters because AI assistants are being designed to cultivate trust. OpenAI’s ChatGPT, Google’s Gemini, and Apple’s Siri are all being positioned as personal helpers—entities that know your preferences, your schedule, your health concerns, and your financial situation. The more personal data these assistants absorb, the more persuasive their recommendations become. And the more persuasive the recommendations, the more valuable the advertising real estate.
Think about what happens when an AI assistant knows that you’ve been searching for running shoes, that your credit card was recently charged at a physical therapy clinic, and that you asked about knee pain last Tuesday. A shoe company willing to pay for placement could have its product recommended not as an ad, but as a thoughtful, personalized suggestion from a trusted digital companion. The line between helpful advice and paid promotion doesn’t just blur—it ceases to exist.
Google’s AI Overviews: The Template for What’s Coming
We’re already seeing early versions of this play out. Google’s AI Overviews, which now appear at the top of many search results, have begun incorporating sponsored content. In May 2025, Google confirmed that ads would appear within AI Overview responses, formatted to look similar to the AI-generated text around them. The company insists these will be clearly labeled, but early user testing suggests that many people don’t distinguish between the AI-generated answer and the embedded advertisement.
According to reporting from multiple technology outlets, Google’s internal projections show AI Overviews as a major growth vector for ad revenue over the next three years. The logic is simple: if users increasingly rely on AI-generated summaries instead of clicking through to websites, then the AI summary itself becomes the prime advertising location. The websites that once hosted the content—and the ads—become irrelevant middlemen.
Amazon’s Alexa and the Commerce Funnel
Amazon’s approach is perhaps even more direct. Alexa, the company’s voice assistant installed in over 500 million devices worldwide, has long been a vehicle for driving purchases on Amazon’s marketplace. With the integration of more advanced large language models into Alexa—a project Amazon has been investing billions in—the assistant becomes capable of more sophisticated product recommendations, comparison shopping, and even anticipatory purchasing.
The Juno Labs piece raises a pointed question about Amazon’s model: when Alexa suggests a specific brand of paper towels, is that recommendation based on your past preferences, on objective quality metrics, or on the fact that the brand is paying for preferred placement in Amazon’s advertising console? The answer, for most consumers, is unknowable. And that opacity is the feature, not the bug.
Apple’s Privacy Promise Faces Its Hardest Test
Apple has long marketed itself as the privacy-first alternative. But even Apple’s business model is shifting. The company’s Services division, which includes the App Store, Apple TV+, iCloud, and increasingly, advertising, generated over $96 billion in revenue in fiscal year 2024. Apple’s search ads business within the App Store is already a multi-billion-dollar operation, and analysts at Morgan Stanley have projected that Apple could build an advertising business worth $30 billion or more by 2027.
With Apple Intelligence now deeply integrated into iOS, the company has the infrastructure to serve contextually aware suggestions across every app and interaction on the iPhone. Apple insists that its on-device processing model protects user privacy, but the Juno Labs analysis notes that privacy and advertising are not inherently incompatible—Apple can serve highly targeted suggestions without ever sending raw data to external servers. The targeting just happens locally, which may actually make it more effective and harder to audit.
Regulatory Frameworks Are Years Behind
Current advertising disclosure regulations were written for a world of banner ads and television commercials. The Federal Trade Commission’s guidelines require that paid endorsements and advertisements be clearly disclosed, but these rules were not designed for a world where a conversational AI agent weaves commercial recommendations into what feels like personal advice. The European Union’s AI Act, which began phased implementation in 2025, includes provisions around transparency for AI systems, but enforcement mechanisms remain underdeveloped.
The fundamental challenge is one of attribution. In a traditional ad, you can point to the sponsor, the placement, and the disclosure. In a conversational AI response, the recommendation is generated by a model trained on data that may include commercial relationships, reinforcement learning from human feedback that may prioritize certain outcomes, and retrieval-augmented generation pipelines that may draw from sponsored content sources. Determining whether a specific recommendation was influenced by advertising becomes a technical and legal puzzle of extraordinary complexity.
What Consumers and Businesses Should Demand
The Juno Labs analysis ends with a call that deserves amplification: consumers should demand the same disclosure standards for AI recommendations that they expect from human financial advisors, doctors, and journalists. If a recommendation is influenced by a commercial relationship, that influence should be disclosed—clearly, consistently, and in a format that the average user can understand.
For businesses, particularly small and mid-sized companies that lack the budgets to pay for AI placement, the stakes are existential. If AI assistants become the primary way consumers discover products and services, and if those assistants prioritize paying advertisers, then the competitive playing field tilts dramatically toward incumbents with deep pockets. The open web, already under pressure from platform dominance, could see its role as a discovery mechanism diminish further.
The companies building AI assistants will protest that their interests are aligned with users—that a helpful assistant is a profitable assistant. But the history of advertising-supported technology tells a different story. From Facebook’s News Feed algorithm to Google’s search rankings, the pattern is consistent: when advertising revenue is the primary business model, the product is optimized for the advertiser first and the user second. There is no reason to believe AI assistants will be any different. The question is whether regulators, journalists, and consumers will recognize this before the most powerful advertising channel ever created becomes the default interface for daily life.