That question you asked ChatGPT about your health symptoms, your finances, or your job search isn’t just a conversation. For users on Free and Go plans, it’s now an secretly tracking users advertising signal.
OpenAI has confirmed a gradual rollout of ads to eligible Free and Go users in selected regions. Paid subscribers on Plus, Pro, Business, Enterprise, and Edu plans do not receive ads.
What Your Prompts Actually Signal
Your current conversation is one of several inputs that can determine which ad you see.
OpenAI’s help documentation says ad selection considers the context and intent of your current conversation, advertiser-supplied topic descriptions, the ad’s own content, and, when personalization is enabled, selected signals from your broader ChatGPT activity. The list also includes the ad’s landing page and advertiser targeting selections, making the full set of inputs broader than it might first appear.
Advertisers supply what OpenAI calls “context hints,” which are descriptions of conversations, topics, or keywords where their product might be relevant. OpenAI says these hints guide matching but do not function as exact-match keywords and do not guarantee delivery in a specific conversation.
Turning off ad personalization narrows the signals used, but it does not remove ads entirely. OpenAI says users who disable personalization may still see ads based on the current chat thread, while other chat threads, ad history, and topics are not used for ad selection.
Your Email Address May Already Be in the System
Advertisers can build custom audiences from identifiers they already hold, before you type a single word.
A 404 Media investigation reported that advertisers can upload customer or prospect lists to build custom audiences for ChatGPT ad campaigns. According to OpenAI’s documentation, supported identifier types include raw email addresses, phone numbers, SHA-256-hashed versions of both, and Google Advertising IDs.
Audience size requirements limit how narrowly that targeting can be applied. OpenAI’s documentation states that inclusion targeting requires at least 25,000 matched users. Smaller audiences may be used for exclusions.
OpenAI states that advertiser-uploaded files are deleted after processing, that advertisers cannot identify which individuals matched their list, and that advertisers cannot select specific users. These are OpenAI’s stated policies, not independently verified facts.
Zach Edwards of DecryptAds, quoted by 404 Media, described OpenAI’s approach as “boiling the frog,” gradually monetizing users while hoping they won’t notice the privacy implications.
The more consequential concern isn’t that ads exist. It’s that an advertiser who already holds your email address from another platform can potentially use that identifier to define a ChatGPT audience. That connection links your existing advertising identity to your conversational behaviorvertisers do not receive your chats or personal details
What OpenAI Says You Can Do About It
Several controls are available, but disabling personalization does not equal opting out of ads.
OpenAI says users can manage ad personalization, dismiss ads, submit feedback, and view an explanation of why a specific ad was shown, all through ad controls in the platform. OpenAI also states that your conversations are not shared with advertisers or marketing partners, and that ads do not influence ChatGPT’s answers.
The Gap Between Policy and Expectation
Formal controls exist, but they only help if you know they’re there.
The transparency issue isn’t whether formal controls exist. It’s whether users who treat ChatGPT as a private thinking space understand it can simultaneously function as an ad-targeting surface. If you use a Free or Go plan, the practical step is to review your ad personalization settings and understand what disabling them does and doesn’t change. Exploring AI-Powered Websites that offer alternative tools may also be worth considering. Your current conversation remains a targeting signal either way, for any user eligible to receive ads.
