Your chatbot conversations can now help decide which ads you see

What you type into an AI chatbot can now help decide which ads you see. Since December 16, 2025, chats with Meta AI help shape content and ads on Facebook and Instagram. ChatGPT began showing ads to users in India in late August 2026. Both moves turn long, personal conversations into a possible source for advertising.

What changed

Old ad tracking worked from clicks and searches. A search for shoes shows interest and little else. A chat can lay out a whole situation, such as a tight budget, a health worry or a new baby. AI can infer intent and personal circumstances from that kind of detail.

Meta says the Meta AI change applies in most regions, except the UK, the EU and South Korea. Users cannot opt out of having their AI chats counted, though ad personalization settings can still be adjusted. Meta also says topics such as religion, health, politics, sexual orientation, and racial or ethnic origin stay out of ad targeting.

ChatGPT ads appear as labeled sponsored placements beneath answers. They are shown to logged-in adults on the Free and Go tiers. OpenAI says ads do not influence answers, advertisers never receive private conversations, and ads will not appear near health, mental health or political topics. The feature now runs in more than 40 countries.

Why it matters

These are company statements about intended use. They are not technical limits on what a system could infer. A University of Washington study found that 82% of respondents rated chatbot conversations as sensitive or highly sensitive. That reflects perception, not proven misuse.

There are three separate questions to ask. Does the platform use a chat to pick an ad? Does the advertiser receive the chat? Does the platform keep it for other purposes? An answer to one says nothing about the other two.

India's legal gap

India's Digital Personal Data Protection Act, 2023, allows personal data to be processed based on consent or specified legitimate uses. Its main obligations and penalties apply from May 2027, after ChatGPT ads arrived. The law does not automatically ban inference-based advertising. Whether a practice complies depends on the provision and how the data moves.

What to do

• Keep full names, addresses, bank details and health information out of chatbots.
• Check memory, chat deletion and ad personalization settings separately.
• Limit app permissions and review the interest profile a platform has built.
• Consider a tracker blocker to reduce casual data collection.