OpenAI deploys ChatGPT Ads in India: 50 brands, WPP and Omnicom as partners, self-serve at 725 rupees/day
🔎 India becomes ChatGPT's second advertising lab
On August 27, 2026, OpenAI officially rolled out sponsored ads in ChatGPT on the Indian market. This is not a discreet test: 50 brands are active from day one, backed by two industry giants — WPP and Omnicom. India was not chosen at random. With over 100 million weekly users, it is ChatGPT's second-largest market in the world, behind the United States.
This Indian rollout comes three months after ChatGPT Ads opened to US advertisers, marking a clear acceleration in OpenAI's monetization strategy. The context is anything but trivial: the company modified its terms of use in August 2026 to allow advertising, and all this while rumors of an IPO are in full swing.
The question is no longer whether ChatGPT will become an advertising medium, but how fast.
The key points
- Targeted ads deployed on ChatGPT's Free and Go plans in India since August 27, 2026, for logged-in adult users.
- 50 brands launched on day one, via agencies WPP and Omnicom as initial partners.
- Self-serve Ads Manager scheduled for September 4, 2026, with a minimum daily budget of 725 rupees (approximately 8 euros).
- 1 in 5 queries is expected to display an ad according to estimates from BigGo Finance.
- Paid plans spared: Plus, Pro, Business, Enterprise, and Education remain strictly ad-free.
- Clear visual separation between ChatGPT's responses and the ads, with clear "Sponsored" labeling.
Recommended tools
| Tool | Main usage | Price (August 2026, check on openai.com) | Ideal for |
|---|---|---|---|
| ChatGPT Ads Manager | Targeted advertising in ChatGPT | Starting at 725 ₹/day (self-serve from Sept. 2026) | Indian advertisers targeting the 100M+ users |
| ChatGPT Free | General public AI assistant | Free | Casual users |
| ChatGPT Pro | Ad-free AI assistant, advanced models | $20/month (check on openai.com) | Pros who refuse advertising |
Who sees the ads and where exactly?
Ads appear only for logged-in adults on ChatGPT's Free and Go plans in India. Not for minors, not for logged-out users, not on paid plans.
Specifically, the ad appears below ChatGPT's responses, in a clearly defined space labeled "Sponsored". OpenAI emphasizes that the model's responses are not influenced by advertisers: the ad is generated independently of the AI response.
Certain topics are excluded. Queries related to health, personal finance, politics, or sensitive content do not trigger ads. This restriction can be compared to the recent announcement of personal finance management for Pro users: OpenAI maintains a strict boundary between highly sensitive features and ad monetization.
The format resembles what you find in the best AI tools for marketing — native, contextual ads, but with an additional promise: conversational relevance. In theory, a query about "best laptop for video editing" could trigger a Lenovo or HP ad directly below ChatGPT's response.
50 brands at launch: WPP and Omnicom at the forefront
The choice of WPP and Omnicom as inaugural partners is not coincidental. They are the two largest communication groups in the world, and their presence from day 1 legitimizes the channel in the eyes of other advertisers.
According to Superpower Daily, 50 brands are active at launch. Brand Equity (Economic Times) confirms that WPP and Omnicom have secured preferred partner positions, likely with early access conditions and guaranteed volumes.
Why India? Three converging reasons.
First, the user volume. India exceeds 100 million weekly users on ChatGPT, an audience pool comparable to the largest Indian digital platforms. Second, the Indian advertising market is growing rapidly — it surpassed $4 billion in 2025 according to industry estimates. Third, the cost-to-audience ratio is attractive: a minimum budget of 725 rupees per day (around 8 euros) provides access to an audience that traditional social networks charge significantly more for at CPM.
The strategy is clear: validate the model in India, then expand it to other emerging markets, and finally generalize.
Ads Manager self-serve: 725 rupees/day starting September 4
This is the real signal from OpenAI. Until September 4, 2026, access is reserved for large advertisers via WPP and Omnicom. After that, any Indian advertiser will be able to create campaigns via a self-serve Ads Manager.
According to PingAura and India Today, the Ads Manager features include:
- Minimum daily budget: 725 rupees (approximately 8 euros, August 2026)
- Currency: Indian rupees (INR)
- Targeting: by conversation topic, demographics, usage behavior
- Format: sponsored ads below responses, with a clickable link
- Reporting: impressions, clicks, CTR, conversions
For comparison, the minimum daily budget on Google Ads in India is around 50 rupees, and on Meta Ads, 35 rupees. OpenAI is clearly positioning itself higher in terms of barriers to entry, suggesting a premium ambition on CPM.
The fact that campaigns are in rupees is a strong signal: OpenAI is not treating India as a secondary market. The billing infrastructure is localized, making it easier for Indian SMBs to adopt.
The business model behind ads on ChatGPT
OpenAI is burning through cash. 2025 estimates put the company's annual losses between $3 and $5 billion, primarily due to compute infrastructure costs. The $20/month ChatGPT Plus subscription is not enough to cover the costs of the most active users — some cost more than they bring in.
Advertising solves a simple equation: Free and Go users are expensive in terms of compute but pay nothing. Monetizing them through advertising turns a cost center into a revenue center, without reducing the perceived quality of the service.
The ratio of 1 ad for 5 queries (20% of interactions) is crucial. It is infrequent enough not to alienate users, but frequent enough to generate significant ad volumes. If ChatGPT generates 10 billion queries per week worldwide, and 20% display an ad with a $15 CPM, we are potentially talking about tens of millions of dollars per month.
This model echoes Google Search's strategy in the 2000s: offering a free and powerful service, funded by non-intrusive contextual advertising. The difference is that ChatGPT already has a base of paying subscribers who indirectly subsidize free users. Advertising reverses this dynamic.
OpenAI's Step-by-Step Monetization Strategy
The Indian rollout is not an isolated event. It is part of a precise sequence.
Step 1 — United States, May 2026. Opening of ChatGPT Ads to US advertisers via agency partners. First real-world test, analyzed in detail here. Reactions from US users were mixed but not massively negative — the tolerance threshold was not crossed.
Step 2 — TOS Modification, August 2026. OpenAI updates its terms of service to explicitly allow advertising in ChatGPT. This is the legal aspect that precedes geographic expansion.
Step 3 — India, August 27, 2026. Rollout in the world's second-largest market, with a self-serve Ads Manager planned a week later. The Indian market serves as large-scale validation before a potential expansion into Europe and Southeast Asia.
Step 4 — IPO. Although OpenAI has not officially announced an initial public offering, its valuation exceeds $300 billion and investors are pushing for an exit. A diversified revenue model with predictable advertising revenue is a strong argument for an IPO.
Each step is designed to progressively normalize the presence of advertising in ChatGPT, without a sudden shock to users.
Free and Go with ads, Pro and beyond protected
Segmentation is strategic and deserves to be analyzed in detail.
| Plan | Ads ? | Logic |
|---|---|---|
| Free | Yes | Direct monetization of a massive non-paying base |
| Go | Yes | Low-priced intermediate plan, ads compensate for the reduced rate |
| Plus | No | The subscriber pays $20/month, ads would drive them away |
| Pro | No | Premium plan at $200/month, ads are incompatible with the positioning |
| Business/Enterprise | No | B2B clients, ads in a professional tool are a commercial non-sense |
| Education | No | Ethical and regulatory sensitivity |
This architecture creates a clear aspiration effect: not seeing ads becomes a selling point for paid plans. This is exactly the model that worked for Spotify, YouTube, and Hulu. The difference is that ChatGPT is primarily a productivity tool, not an entertainment one. Tolerance for advertising there is potentially lower.
The Go plan, recently launched as an intermediate option, finds itself in an uncomfortable position: its users pay a small amount but still see ads. If the price difference between Go and Plus is not perceived as justified, this could create a friction effect.
What Indian advertisers gain (and risk)
For Indian brands, ChatGPT Ads represents an unprecedented opportunity: reaching users in a moment of high intent. When someone asks ChatGPT "what is the best phone under 15,000 rupees", the query expresses a purchase intent that neither Google Search nor social networks capture with the same contextual richness.
However, the risks exist. The first is the proximity between the AI response and the ad. Even with clear labeling, some users might associate ChatGPT's recommendation with the sponsored ad, which would create confusion detrimental to both the advertiser and OpenAI.
The second risk is the perceived quality of the model. If a user asks a serious question and ends up with an ad for a mobile game under their response, the experience is degraded. OpenAI states that sensitive topics are excluded, but the definition of "sensitive" remains vague.
The third risk is measurable: the click-through rate. Ads under the responses of an LLM like GPT-5.5 or Gemini 3.1 Pro do not benefit from the clicking habit forged by years of Google Ads. ChatGPT users are in a conversational mode, not in a commercial search mode. The initial CTR could be disappointing.
The competition: how Google and Meta are reacting
The Indian advertising market is dominated by Google and Meta, which together control over 70% of digital spending. OpenAI's entry into this space is a direct provocation.
Google is particularly threatened. ChatGPT Ads targets the exact same moment of intent as Google Search: the user has a problem, they formulate it in natural language, they expect a useful response. The difference is that ChatGPT provides a synthesized response instead of a list of blue links. If Indian advertisers find that the ROAS on ChatGPT is competitive with Google Search, a budget shift is conceivable.
Meta, for its part, is less directly threatened because its model relies on demographic and behavioral targeting, not on search intent. But the growth of advertising budgets is not infinite: every rupee spent on ChatGPT is a rupee that does not go to Instagram or Facebook.
Moreover, the current AI trends show that major advertising groups are starting to allocate dedicated budgets to generative AI channels, in addition to traditional channels. WPP and Omnicom understood before the others that this channel will exist — they might as well be there from day one.
Safeguards: Independence of responses and privacy
OpenAI published an official post "Testing ads in ChatGPT" detailing the safeguards surrounding this rollout. Three points deserve attention.
The independence of responses. OpenAI asserts that the generative model does not know which ad will be displayed when it produces its response. The ad is injected after generation, in a separate layer. This is a crucial point for credibility: if ChatGPT's responses were influenced by advertisers, trust in the product would collapse.
Clear labeling. Each ad carries the "Sponsored" label and is visually separated from the response. OpenAI has an interest in making this separation as obvious as possible, as any ambiguity would backfire on them.
Data protection. OpenAI specifies that conversations are not used for ad targeting on an individual basis. Targeting relies on thematic categories (the topic of the conversation) rather than on the specific content of the exchanges. This is an important distinction from a regulatory perspective, especially in India where the Personal Data Protection Act (DPDPA) is gradually coming into effect.
These safeguards are necessary but not sufficient. Trust is built over time, and a single incident of contamination between a response and an ad could call everything into question.
The broader context: AI, advertising, and ethics
Deploying advertising within an AI assistant raises questions that go beyond the OpenAI case.
When Claude Opus 4.7 or GPT-5.5 answer a medical question, even the slightest appearance of a conflict of interest would be disastrous. OpenAI has excluded health topics from ad targeting, but the boundary between health and wellness is porous. Is a query about "sleep supplements" a health topic or a lifestyle one?
OpenAI's Rosalind project, which turns ChatGPT into a biodefense agent, shows just how much the company is deploying ChatGPT in highly critical fields. The coexistence between these uses and advertisements for consumer brands creates a strong identity tension.
Anthropic, with Claude, has taken a stand against advertising in its assistant. Google inserts commercial suggestions into Gemini in a more subtle way. OpenAI is choosing the most aggressive approach, and India is the most favorable testing ground to validate this choice.
❌ Common mistakes
Mistake 1: Thinking ChatGPT Ads replaces Google Ads in India
This is not the case today. The impression volume on ChatGPT is minuscule compared to Google Search in India (billions of queries per day). ChatGPT Ads is a complementary channel, not a substitute. Advertisers who go into it thinking they will replace Google will be disappointed by the volumes.
Mistake 2: Aiming for a CPM similar to social networks
ChatGPT audiences are qualitatively different from Facebook or Instagram audiences. The user is in problem-solving mode, not passive scrolling mode. The CPM will likely be higher, but so will the CTR and potential conversion. Comparing them directly is an analytical error.
Mistake 3: Ignoring segmentation by plan
Not all ChatGPT users are equal for an advertiser. Free and Go users have different profiles — Go users have shown a willingness to pay (even a small amount), which is a value signal. Rough targeting that mixes Free and Go loses precision.
Mistake 4: Believing that responses are influenced by ads
Even if OpenAI claims independence, some advertisers might hope for a halo effect ("ChatGPT talked about my industry, so the ad will be more effective"). This effect, if it exists, is unmeasurable and not guaranteed. Building a strategy on it is hazardous.
❓ Frequently Asked Questions
Who sees ads in ChatGPT in India?
Only logged-in adult users on the Free and Go plans. Minors, logged-out users, and all paid plans (Plus, Pro, Business, Enterprise, Education) do not see them.
What is the minimum budget to advertise on ChatGPT in India?
725 rupees per day (approximately 8 euros, August 2026) starting from the launch of the self-serve Ads Manager on September 4, 2026. Before this date, access is exclusively through WPP and Omnicom.
Are ChatGPT's answers influenced by advertisers?
No, according to OpenAI. Answers are generated independently, and ads are injected afterwards in a separate layer. Targeting is based on the conversation topic, not by manipulating the generated content.
Why India as the second market?
India has over 100 million weekly ChatGPT users, making it the second-largest market globally. The advertising market there is growing rapidly, and ad acquisition costs are attractive to validate the model before a wider expansion.
Does this rollout signal an OpenAI IPO?
Nothing is official, but revenue diversification (subscriptions + API + advertising) is a classic prerequisite for an IPO. The timing — changes to the T&Cs, ad rollout, revenue growth — is consistent with IPO preparation.
✅ Conclusion
The deployment of ChatGPT Ads in India marks OpenAI's transition from AI startup to media player. With 50 brands at launch, an Ads Manager at 725 rupees/day, and the protection of paid plans, the strategy is calibrated to maximize revenue without killing the user experience. It remains to be seen whether advertisers will achieve the expected ROAS — and whether Indian users will accept ads in their daily AI assistant. To follow the evolution of this model and compare the approaches of different players, check out our Claude vs ChatGPT comparison and our selection of the best AI tools.