The AI Ads Split: What OpenAI's $1 Billion Bet and Anthropic's Anti-Ad Campaign Mean for Your Business

If you use ChatGPT, Perplexity, Claude, or Copilot at work, the company behind it has placed a very public bet on how it plans to make money from you — and the biggest AI labs have made opposite bets. In late August, OpenAI announced its ChatGPT advertising business had crossed a $1 billion annualized revenue run rate in under 200 days — faster than Google’s AdWords took to reach the same milestone. That milestone lands the same year Perplexity announced it was winding down its advertising program entirely, citing user trust, and Anthropic spent millions on a Super Bowl campaign built around the tagline “Ads are coming to AI. But not to Claude.” Microsoft’s Copilot, meanwhile, has been quietly rolling ads into its answers, drawing a different kind of headline: user backlash over broken promises.

This isn’t a marketing curiosity. It’s a monetization fork in the road for tools enterprises are wiring into sales pipelines, support desks, and internal workflows, with real implications for data handling and vendor trust — the same category of question companies already ask when they run a 10-minute AI risk audit. This post breaks down what each company did, why they’re moving in opposite directions, and what IT and procurement teams should ask before signing or renewing any AI contract.

Three Companies, Three Opposite Bets

OpenAI: Ads as a Second Growth Engine

OpenAI’s ad business isn’t a side experiment. According to OpenAI’s own announcement and reporting from CNBC, the ChatGPT ads product — shown to users on the Free and Go tiers — reached a $1 billion annualized run rate roughly six months after launch, a pace Forbes notes beat Google’s four-year climb to the same number with AdWords. OpenAI frames this as proof of a “diversified business model” as it prepares for what’s expected to be a massive IPO at a reported $852 billion valuation.

Crucially, Plus, Pro, Business, Enterprise, and Edu accounts remain ad-free — only the free tier carries ads. That distinction matters more than it looks, and we’ll come back to it.

Perplexity: Betting on Trust Over Ad Revenue

Perplexity took the opposite path. After launching an ad-supported publisher revenue-share program in late 2024, it announced in early 2026 that it was phasing out ads entirely by year’s end. Executives told the Financial Times, as reported by Campaign US, that a user has to believe an answer is genuinely the best one, not a sponsored one, to keep trusting and paying for the product. The bet paid off fast: eMarketer reports Perplexity’s monthly recurring revenue jumped roughly 50% the following month, pushing annualized revenue past $450 million, driven entirely by subscriptions and enterprise licenses.

Anthropic: Turning “No Ads” Into a Marketing Weapon

Anthropic went further, building an entire Super Bowl campaign around the decision. The spots — “Deception,” “Betrayal,” “Treachery,” and “Violation” — all closed with “Ads are coming to AI. But not to Claude,” a direct shot at OpenAI reported by CNBC and Fortune. It worked: CNBC later reported Claude’s daily active users rose 11% and site visits jumped 6.5% in the following weeks — enough that OpenAI CEO Sam Altman publicly said he found the ads funny while calling them “so clearly dishonest.”

Google, meanwhile, sits in the middle: it already runs ads inside AI Overviews and AI Mode in Search but, per eMarketer’s analysis, has so far kept them out of the Gemini chat interface itself.

Why This Split Matters Beyond Marketing

The Free-Tier Trap

Ads are confined to free and low-cost tiers, not the paid plans most companies deploy internally. But that’s exactly the trap: employees experimenting with AI tools rarely start on the enterprise tier. They start on the free version on a personal laptop, get hooked, and it quietly becomes part of a workflow nobody signed off on — the same shadow AI pattern security teams already warn about. An ad-monetized free tier means employee prompts are, by design, part of an advertising pipeline the moment that shadow usage happens — a different risk profile than a sanctioned contract with audit logs and key management.

Enterprise Tiers Aren’t Automatically Immune

Even where a vendor promises “no ads on paid plans,” that’s a policy choice, not an architectural guarantee — and policies change as monetization pressure builds. Microsoft is the clearest cautionary example: it began testing sponsored responses woven into Copilot’s actual generated answers, expanding through 2025 and 2026 with features like “Showroom Ads” and branded agents, per eMarketer. The backlash was immediate — WebProNews reported users felt blindsided because the line between a genuine answer and a paid placement had blurred. For a company running Copilot inside Word, Excel, and Teams, that’s a live question about whether AI-generated business content is nudged by advertiser incentives.

This is the governance blind spot covered in The Agent You Never Approved: the risk isn’t always the AI tool you deployed on purpose — it’s the monetization model shifting underneath a tool you already trust.

The Broader Pattern: AI’s Monetization Reckoning

Zoom out and this is a referendum on how AI labs think long-term revenue gets built. OpenAI treats ads as a scalable second engine, useful as it heads toward an IPO that needs to justify an enormous valuation. Perplexity and Anthropic are betting trust is the scarcer resource, worth more in subscription and enterprise revenue than it would ever make in ad dollars. Neither bet looks wrong yet — both showed real revenue growth right after their announcements. But it means the AI vendor landscape enterprises choose from no longer monetizes uniformly. A year ago, nearly every serious AI product was subscription-first; today the same category contains an ad-supported model, an ad-free one, and everything in between.

What Procurement and IT Teams Should Actually Do

None of this means ripping out AI tools. It means treating “how does this vendor make money” as a standing procurement question, the same way teams already ask about data residency when evaluating private, on-device AI processing.

Ask every AI vendor these questions

Which tiers carry ads today, and is that commitment contractual or just current policy?
Is ad-eligibility tied to account tier, or to an opt-out setting a user could accidentally leave unchecked?
If the vendor changes its ad policy, does your contract guarantee notice, or could a Copilot-style rollout arrive unannounced?
For free tools employees adopt informally, does IT know which of them are ad-monetized versions of products you already license correctly elsewhere?

Add a line to procurement checklists and renewal reviews that explicitly asks about advertising and monetization plans, not just data training and retention — most AI risk frameworks were built around data leakage, not around whether an answer is quietly shaped by an advertiser relationship.

Conclusion

This year made something clear: there is no single “AI business model” anymore. OpenAI is scaling ads as fast as any ad platform in history. Perplexity walked away from ads and grew faster because of it. Anthropic turned the absence of ads into its most effective marketing campaign to date. Microsoft’s rollout shows what happens when the shift isn’t communicated well. For any business building workflows on top of these tools, the practical takeaway isn’t to pick a side in the ad debate — it’s to stop assuming your AI vendor’s monetization model is static, and start asking the question explicitly, on every contract, at every renewal, for every tool your teams actually use day to day.

Frequently Asked Questions

Does ChatGPT show ads to paying business customers?

No. OpenAI has stated that Plus, Pro, Business, Enterprise, and Edu accounts remain ad-free; ads currently only appear for users on the Free and Go tiers.

Why did Perplexity stop showing ads if it needed revenue?

Perplexity’s leadership concluded that ads undermined user trust in its answers, which it saw as more valuable long-term than ad revenue. After phasing ads out, the company reported roughly 50% month-over-month growth in recurring revenue, reaching over $450 million annualized, driven by subscriptions and enterprise licenses.

Is Microsoft Copilot’s advertising the same as ChatGPT’s?

Not exactly. Copilot’s ads have appeared as sponsored content and branded agent responses woven into generated answers, which drew criticism for blurring the line between genuine advice and paid placement. ChatGPT’s ads on free tiers are presented as separate, clearly labeled units rather than embedded in the answer text.

Should my company avoid free-tier AI tools because of this?

Not necessarily, but IT and security teams should know which AI tools employees are using informally, on which tier, and whether that tier is ad-supported — the same due diligence already applied to shadow SaaS tools.

Could an “ad-free enterprise tier” promise change in the future?

Yes. A no-ads policy on a paid tier is a business decision, not a fixed technical guarantee, and can change as a company’s monetization pressure shifts — which is why it’s worth confirming contractually rather than assuming it’s permanent.

Are other major AI companies planning to add ads?

Google already places ads in AI Overviews and AI Mode in Search but has so far kept them out of the Gemini chat interface itself. Anthropic has publicly committed to keeping Claude ad-free as part of its brand positioning.