The arrival of ads in ChatGPT reads like one company’s decision. The more interesting fact is that the other AI companies answered the same question in completely different ways. OpenAI turned advertising into a major revenue line. Anthropic committed in writing to staying ad-free. Perplexity tried ads and reversed. And Google publicly denied a report that ads were coming to its Gemini app.
This post maps the landscape using verifiable announcements only.
Summary Table
| Platform | Ad status | Basis |
|---|---|---|
| ChatGPT | Yes. Live since February 2026, on Free and Go plans | OpenAI announcements and help center |
| Google AI Mode | Yes. Formats positioned inside the answer | Google Marketing Live, May 20, 2026 |
| Google Gemini app | No. Report officially denied | Dan Taylor (Google), December 8, 2025 |
| Meta AI | Indirect. Conversations feed ad personalization | Meta announcement, October 1, 2025 |
| Claude | No. Written ad-free commitment | Anthropic, February 4, 2026 |
| Perplexity | Removed. Tried it, reversed | February 2026 reporting |
ChatGPT: Turning Ads Into a Revenue Line
OpenAI took ads from a US test on February 9, 2026 to more than fifty markets in seven months, and its August 31, 2026 announcement reported passing $1 billion in annualized revenue run rate in under 200 days. Ads appear only on the Free and Go plans; Plus, Pro, Business, Enterprise and Edu remain ad-free.
How the channel works is covered in our ChatGPT Ads roundup, and what advertisers actually got from it in the published campaign results.
Claude: Putting Ad-Free in Writing
In a statement dated February 4, 2026, Anthropic set out its position plainly: Claude will remain ad-free, users will not see sponsored links beside their conversations, Claude’s responses will not be influenced by advertisers, and they will not include third-party product placements the user did not ask for.
Two details make the commitment more interesting than a slogan.
First, the door is not sealed. The same text says that if the company needs to revisit the approach, it will be transparent about its reasons. This is a reasoned choice, not a “never”.
Second, ad-free does not mean commerce-free. Anthropic says it is enthusiastic about agentic commerce, where Claude handles a purchase or booking end to end on a user’s behalf. The line it draws is about who starts the interaction: the user initiates, not the advertiser.
That distinction matters more than it first appears. An agent completing a purchase and an advertiser pushing a product into an agent are technically adjacent and commercially opposite.
Perplexity: The One That Tried and Reversed
This is the most instructive case. Perplexity ran ads and pulled them. In February 2026 it was reported to be removing the sponsored answers that appeared beneath responses, having phased advertising out over the previous year, with no plan to return.
The stated reason was not technical but about trust. The company’s framing pointed to the risk that ads make a user start doubting everything. The problem was not the ad unit; it was placing an ad in the same interface as a claim to accuracy.
For advertisers this is not a trivia item. An AI assistant’s answer is valuable precisely while it is perceived as impartial. When that perception erodes, both organic and paid visibility lose force.
Google: Two Products, Two Different Answers
Google is where the confusion — and the rumour — lives, because two separate products get treated as one.
AI Mode, inside Google Search: ads exist. Google announced formats positioned inside the generated answer at Marketing Live 2026.
The Gemini app, the standalone assistant: Google denied it. Adweek reported that Google representatives told at least two advertising clients that ad placements in Gemini were targeted for a 2026 rollout. The same day, Google’s VP of Global Ads Dan Taylor responded that the story rested on uninformed, anonymous sources, stating there are no ads in the Gemini app and no current plans to change that.
Reading this correctly means separating a briefing claim from a company’s official position. They contradict each other, and as of today the official position is that there are no ads. Media plans should rest on that statement rather than on the briefing rumour.
Meta AI: Not Ads, but Ad Data
Meta’s approach is structurally different from the rest. Rather than placing ads inside Meta AI, it turns conversations with the assistant into an input for the existing ads system.
Per Meta’s October 1, 2025 announcement, interactions with Meta AI have been used to personalize content and ad recommendations since December 16, 2025. Users were notified in advance through in-product notifications and email, and Meta says it uses this information across the accounts a person adds to the same Accounts Center.
One boundary is drawn explicitly: conversations about topics such as religion, sexual orientation, political views, health, racial or ethnic origin, philosophical beliefs and trade union membership are not used to show ads.
For advertisers the implication is that there is no new inventory to buy on Meta’s side. The targeting signal behind existing inventory simply got richer, which is why the first-party data and consent conversation sits on different ground here.
What This Means for Media Planning
Three practical conclusions.
An “AI assistant” is not one channel. Inventory can currently be bought in ChatGPT and Google AI Mode. In Claude and Perplexity there is nothing to buy, so visibility there is purely organic — the GEO and AEO side of the work.
Ad-free platforms are not gaps in the plan. A user can ask Claude and later see your ad in ChatGPT, or the reverse. Being cited as a source in an ad-free assistant produces a gain that never appears in any paid metric.
Two business models are competing in the open. OpenAI produced fast revenue from ads. Anthropic and Perplexity converted the absence of ads into a trust argument. No winner is evident yet, and that uncertainty is reason enough to treat channel decisions as experiments rather than long-term commitments.
What We Could Not Verify
For transparency: while preparing this piece we encountered claims about Microsoft Copilot, xAI Grok, Amazon Rufus, Apple and Mistral. Some appeared only on sites whose business is selling AI advertising services, and others could not be traced to an official announcement. No figure or date we could not verify made it into this post. We will update it when those platforms produce something confirmable from an official source.
Related News
- ChatGPT Ads Opens in Türkiye
- What Is ChatGPT Ads and How Does It Work?
- What Advertisers Actually Got From ChatGPT Ads
- How to Set Up Conversion Measurement on ChatGPT Ads
- GEO and AEO: AI Search Optimization
Sources
- Anthropic: Claude is a space to think (February 4, 2026)
- PYMNTS: Perplexity Pulling Sponsored Answers From AI Platform (February 18, 2026)
- Search Engine Land: Perplexity stops testing advertising
- PPC Land: Google denies Gemini ads despite advertiser briefings on 2026 rollout (December 8, 2025)
- Search Engine Land: Google denies ads are coming to Gemini in 2026
- Meta: Improving Your Recommendations on Our Apps With AI at Meta (October 1, 2025)
- Google: A new generation of ads for the AI era of Search (May 20, 2026)
- OpenAI: A milestone in expanding access to AI (August 31, 2026)
Frequently asked questions
- Which AI assistants carry advertising?
- Ads have been live in ChatGPT since February 2026, shown to Free and Go plan users. Google announced ad formats inside AI Mode. Meta began using Meta AI interactions to personalize content and ad recommendations from December 2025.
- Does Claude have ads?
- No. In a February 4, 2026 statement, Anthropic wrote that Claude will remain ad-free, that users will not see sponsored links beside their conversations, and that Claude's responses will not be influenced by advertisers. It added that if it ever revisits the approach, it will be transparent about why.
- Does Perplexity still sell ads?
- No. In February 2026 it was reported to be removing the sponsored answers that appeared beneath responses. The company had been phasing advertising out over the previous year and does not plan to return to it.
- Are ads coming to the Gemini app?
- Google officially denied it. Adweek reported that Google told advertising clients ads were targeted for the Gemini app in 2026. The same day, Google VP of Global Ads Dan Taylor said there are no ads in the Gemini app and no current plans to change that. This is separate from ads in AI Mode inside Google Search.
- Is staying ad-free sustainable for an AI assistant?
- Two bets are running in parallel. OpenAI reached a $1 billion annualized revenue run rate on advertising in under 200 days. Anthropic and Perplexity rely on subscription and enterprise revenue and position the absence of ads as a trust argument. It is too early to call a winner.
- Does Meta use Meta AI conversations for ad targeting?
- Per Meta's October 1, 2025 announcement, from December 16, 2025 interactions with Meta AI are used to personalize content and ad recommendations. Conversations touching sensitive topics — religion, sexual orientation, political views, health, racial or ethnic origin, philosophical beliefs and trade union membership — are not used to show ads.