ChatGPT Ads in India Put Local Advertisers in Charge

OpenAI has begun showing ads on ChatGPT in India for free users and the ₹399 ($4) Go plan. The company says the tiles fund cheaper access.

On August 31 it said ChatGPT Ads had already hit a $1 billion annualized run rate, and that self-serve buying was opening in India that day.

The Ad Goes Below the Answer

Logged-in adults on Free and Go now see a sponsored product or service under a reply when OpenAI finds a match. The company says the tile is labeled, set apart from the answer, and barred from changing what ChatGPT writes. “Answer independence is non-negotiable,” an OpenAI spokesperson said.

Dave Dugan, vice president and head of global ads solutions, is selling the slot as a moment of intent, not a banner on a feed. People in India, he said, already use the chatbot to ask, compare, and weigh tradeoffs, so a brand can show up while a choice is still forming.

India is one of the most dynamic markets in the world for how people use ChatGPT to learn, build, and discover. Our weekly active users have doubled year-on-year, with many of them going beyond simple searches and changing how they make decisions.

Dave Dugan, Vice President and Head of Global Ads Solutions, OpenAI

He also said OpenAI expects ads on about one in five ChatGPT queries in India, and none on the rest, which lack a commercial hook. That is the product in one ratio: most chats stay clean, and the paid tile fires when someone is already in the market.

Users can dismiss an ad, ask why it appeared, manage personalization, or wipe the data used to target them. Free accounts can drop ads in exchange for fewer daily messages. Plus, Pro, Business, Enterprise, and Education stay ad-free.

100 Million Weekly Users on Cheap Plans

India is one of ChatGPT’s largest and most active markets, OpenAI has said, with more than 100 million weekly active users as of February. Dugan said that weekly base has doubled from a year earlier. A crowd that size, on a $4 plan or on no plan at all, is expensive to serve and hard to convert at U.S. prices.

ChatGPT Go arrived in India in August 2025 as a local, low-cost tier. From November 4, 2025, OpenAI made Go free for a year for eligible users around its first DevDay Exchange in Bengaluru. The company also bought attention the old way, spending on cricket around the Indian Premier League and the Women’s Premier League. The cheap plan and the free year built habit. The ads are how that habit starts to pay.

THE INDIA AUDIENCE OPENAI IS NOW SELLING

• Weekly users: More than 100 million as of February, with Dugan saying the figure has doubled year on year.
• Go price: ₹399 a month for higher limits, after a year-long free promo that began on November 4, 2025.
• Opening brands: More than 50 names in the first week, with WPP and Omnicom as the first agency partners.
• Buyer floor: Daily campaign budgets from ₹725, about $7.60, once self-serve is in a shop’s hands.

WPP and Omnicom are on a non-exclusive deal, and OpenAI said it will add more agencies. A spokesperson said early demand should show up in travel, financial services, retail, and health and beauty. That wish list sits next to a stricter global rulebook, which still treats finance and health ads outside the United States as largely off limits.

WHICH INDIA PLANS SHOW CHATGPT ADS

The split is the business. People who already pay more keep a clean chat. People on the plans OpenAI used to grow India become the inventory.

ChatGPT Ads Already Run at $1 Billion

The India user-facing launch on August 27 was not the start of the ads shop. OpenAI began a U.S. test on February 9 for logged-in adults on Free and Go. A March update claimed no hit to consumer trust scores and low dismissal rates. Ads Manager, the self-serve desk, arrived in beta in May.

By August 11 the company had switched the product on in the United Kingdom, Mexico, Brazil, Japan, and South Korea. Twenty days later it said ChatGPT Ads had reached $1 billion in annualized revenue in less than 200 days, with tens of thousands of advertisers and a live footprint in more than 40 countries.

HOW THE ADS PROGRAM GOT TO INDIA

1. February 9, 2026: Opens a U.S. test of ads on Free and Go for logged-in adults.
2. May 5, 2026: Rolls out a beta Ads Manager so brands can buy without a sales team.
3. August 11, 2026: Turns ads on in the United Kingdom, Mexico, Brazil, Japan, and South Korea.
4. August 27, 2026: Starts showing ads to Free and Go users in India, with more than 50 brands in the first wave.
5. August 31, 2026: Says the ads unit is at a $1 billion run rate and that self-serve buying is opening across India, Europe, the Middle East, and North Africa.

OpenAI also cited an ecommerce advertiser at a 3x return on ad spend over 28 days, and a technology partner that said more than 80 percent of ad-driven ChatGPT visits were new customers. Those are vendor numbers, and they are the ones the company wants on a 2027 listing document next to an $852 billion private mark.

In-app ads are already a $400 billion market. ChatGPT is late to that pool, but it is arriving with a chat log instead of a keyword. The India book is how you take a run rate built in richer countries and point it at a huge, cheap audience.

Why Claude Can Stay Ad-Free

Anthropic drew this line in February, five days before OpenAI’s U.S. ads test. In a post titled “Claude is a space to think,” it said a chat with Claude is the wrong place for a pitch, and that Claude will remain ad-free.

There are many good places for advertising. A conversation with Claude is not one of them.

Anthropic, February 4, 2026 statement

Claude is built to be a genuinely helpful assistant for work and for deep thinking.

Advertising would be incompatible with that vision.

Read why Claude will remain ad-free: https://t.co/Dr8FOJxINC

— Claude (@claudeai) February 4, 2026

The lab’s argument is about incentives. A user who mentions bad sleep should get causes and habits, not a path to a sale. Even a labeled tile under the answer, Anthropic wrote, would pull the product toward time-on-app instead of a short, useful reply. It generates money from enterprise contracts and paid plans, and it said it can live with the tradeoffs of skipping ads.

That stance is easier if your customers are companies and developers. OpenAI’s problem in India is the opposite: a mass of people on Free and Go. Anthropic also left itself an exit, saying it would explain itself if it ever reversed course. For now the contrast is the listing story as much as the product story, and it sits beside Anthropic’s public listing pitch as both labs head toward public markets.

The Guardrails Around Those Sponsored Tiles

OpenAI’s public line is that advertisers never see chats, history, memories, or personal details, and that they get only campaign totals such as impressions and clicks. Matching still uses the topic of the live conversation, past chats, and past ad clicks, the company said in its February test note. A recipe thread can pull a meal-kit tile. The advertiser is not supposed to read the thread.

The written rules for ChatGPT ad placement, updated August 31, keep ads off a long list of unsafe or raw contexts, from child safety and self-harm to politics and “emotionally reliant” chats. Early categories are household goods, local services, travel, and digital products or education. Health and finance ads may be approved case by case in the United States and are generally barred elsewhere.

WHERE THE ADS ARE SUPPOSED TO STOP

• Minors: No ads if a user says they are under 18, or if the system predicts they are.
• Paid seats: Plus, Pro, Business, Enterprise, and Education stay clear of the tiles.
• Raw topics: No placements next to mental health, politics, or other sensitive and regulated subjects.
• Chat text: Advertisers get aggregate results, not the words a person typed.
• Look-alike UI: Ads that mimic ChatGPT’s own voice or layout can be pulled.

Interface imitation is banned for a reason. If a sponsored line looks like the model talking, the independence pledge is theater. Plenty of people still say they have not seen a ChatGPT ad at all, which fits a system that only fires on commercial intent and only on two cheap tiers. The open worry is not the label on day one. It is whether a shop that now books real ad money keeps the wall up once quarterly targets exist.

Self-Serve Opened Early for Indian Buyers

OpenAI’s India launch copy pointed to September 4 for Ads Manager. On August 31 the company said self-serve access was opening that same day across India, Europe, the Middle East, and North Africa. By September 2, Indian businesses were posting access emails and working inside the tool.

The original date is still two days away. The product is already in buyers’ hands. That gap is how a global ads desk moves when it is chasing a run rate, not a ribbon-cutting.

The first screen asks for a website and then drafts a preview, an audience, and a plan. In at least one live test on an Indian bakery site, the tool wrote the ad and then recommended a United States location and a $100 daily budget. The desk is open. The defaults still look like they were trained on richer markets. Shops that want a rupee campaign will have to overwrite that map.

That is the opening for local advertisers buying the inventory without waiting on WPP or Omnicom. Managed sales still sit with the agencies and OpenAI’s ads team. Self-serve is how a travel agent or a retailer with ₹725 a day gets in. Households on the other side of the tile already have a clean way out: pay for Plus, or push relatives onto it, the way families once paid to silence music ads.

Google Still Owns the Consumer Side

Dan Niles, founder and portfolio manager at Niles Investment Management, put OpenAI in a bind back in June. It gets squeezed, he said, between Anthropic on enterprise customers and Google, “the king in consumer,” which has the cash to fund a long fight. ChatGPT ads are the attempt to turn a consumer lead into money Google already knows how to collect: a query, an intent, a click.

The India launch is the hard version of that attempt. The users arrived on a ₹399 plan and a free year. The inference bill did not. Ads under the answer are how OpenAI tries to keep those tiers alive without pretending a mass-market chat app can be funded on subscriptions alone. Anthropic can call that a conflict. OpenAI calls it access.

By September 2, Indian businesses were already inside ChatGPT Ads Manager, two days before the date on the original India launch note.