ChatGPT Ad Targeting: Your Prompts, Phone Number, and Email Are Being Used for Precision Advertising

From Conversation to Advertising: ChatGPT's New Commercial Frontier

According to a discussion that gained traction on Hacker News, OpenAI's ChatGPT is using users' prompts, phone numbers, and email addresses to target advertising. This information was disclosed by community users and quickly became a focal point of attention in the tech community.

This means that when users engage in seemingly private conversations with ChatGPT, the prompts used to generate responses—along with the contact information linked to their accounts—may simultaneously be used for commercial ad matching and delivery. For an AI assistant that has become deeply integrated into daily work and life, this shift in data usage touches on critical issues of user trust and privacy boundaries.

What Data Does Ad Targeting Specifically Involve?

According to the discussion on Hacker News, ChatGPT's advertising system combines three types of information to build user profiles and match ads.

Prompt content—This is the most core and most sensitive part. In their interactions with ChatGPT, users may reveal work needs, shopping intentions, health concerns, learning plans, and more. This text information inherently carries extremely high commercial value and can be used to infer users' interests, spending power, and potential needs.

Phone numbers and email addresses—The contact information linked to accounts are themselves strong identifiers. Combined with prompt content, advertisers or the platform can more precisely target specific individuals, and even perform cross-platform user identification.

This combination of multi-dimensional data makes ChatGPT's ad targeting capabilities far exceed traditional delivery methods based on cookies or browsing history. The intent signals captured by conversational AI are superior in both depth and real-time nature.

Strong identifiers is a technical term in the privacy and ad tech field, referring to data fields that can stably and uniquely point to a specific natural person, such as phone numbers, email addresses, ID numbers, or device unique identifiers. In contrast, cookies, browsing history, and similar data are weak identifiers, which may become invalid due to cache clearing or device changes. The core risk of strong identifiers lies in this: once a platform simultaneously possesses strong identifiers and behavioral data (such as conversation content), it can stitch anonymous behavioral traces to real identities, forming associations that are difficult to undo. This is also why phone numbers and emails are more sensitive than prompts themselves in the context of ad targeting—they turn "someone asked something" into "this specific person asked something," and may be used for cross-platform identity matching and remarketing.

Why Is Ad Targeting in Conversational AI More Worrisome?

Traditional search advertising is based on keyword bidding, while social media advertising relies on interest tags and behavioral data. What makes conversational AI like ChatGPT unique is that users often come with clear purposes or genuine confusion, and the conversation content naturally contains abundant context and intent clues.

When users ask ChatGPT "laptop recommendations under 5,000 yuan" or "how to relieve migraines," if this information is used for ad targeting, both its precision and intrusiveness increase significantly. Users may not realize that the process of seeking help from AI is simultaneously a process of actively exposing their needs to the advertising system.

More critically, the privacy of conversation gives people a sense of security. Users tend to view ChatGPT as a neutral tool or assistant, rather than an advertising platform. When this expectation is broken, repairing trust will be extremely difficult.

Intent signals refer to information that users actively express, pointing to clear action goals, such as "want to buy," "want to solve," or "currently comparing." In advertising technology, the value of intent signals far exceeds interest tags because they directly correspond to the critical point of purchase or decision-making. Traditional advertising relies on inference: guessing what you might be interested in through browsing history; while conversational AI captures need statements written by users themselves, even including constraints like budget, symptoms, and time limits. This shifts ad matching from "guessing what you like" to "delivering on demand," with higher conversion efficiency but also greater intrusiveness—users think they are seeking help, but in reality they are completing an unconscious self-profiling annotation.

Industry Trends and Competitive Pressure

ChatGPT's exploration of ad monetization is not an isolated event. The high operating costs of AI assistants—computing power, bandwidth, model iteration—put pressure on relying solely on subscription revenue. Introducing advertising is a natural choice for many free or low-cost AI products seeking a commercial closed loop.

But competitor dynamics may affect the pace of this strategy. If some competitors insist on an ad-free subscription model and make privacy protection a core selling point, ChatGPT's advertising strategy could become a trigger for user migration. The discussion on Hacker News also reflects this concern: the tech community is highly sensitive to data usage, and negative feedback could affect the product's long-term reputation.

The tension between subscription and advertising models is a key backdrop for understanding this round of AI commercialization. The inference cost of large models grows linearly with user scale, while subscription revenue is limited by willingness to pay and conversion rates, making "free + advertising" a compromise that has repeatedly appeared throughout internet history. But AI assistants have an essential difference from search engines and social platforms: users' interactions with them are task-oriented deep conversations, rather than passive browsing. This means ads can be embedded in answers, recommendations, or follow-up questions, in a more native form with blurrier boundaries. Once ads become difficult to distinguish from model output, users' trust in a "neutral assistant" will be eroded—this is the deeper reason why the tech community has reacted so strongly to ChatGPT's advertising strategy.

What Can Users and Developers Do?

Faced with this change, users can take some pragmatic countermeasures. Regularly checking and clearing ChatGPT conversation history, avoiding revealing unnecessary personal identity information in conversations, and considering using separate email addresses and phone numbers to register for non-core services are all ways to reduce data exposure.

For developers, this incident also reminds us that when building applications based on large models, data collection boundaries and user informed consent mechanisms need to be taken more seriously. Behind every API call, there may be end users' privacy expectations involved.

ChatGPT's ad targeting strategy is essentially a repricing between privacy and convenience in the process of AI commercialization. Whether users are willing to exchange commercial use rights to their conversation data for lower direct payments will determine the ultimate direction of this model.

One-line Summary

ChatGPT is reportedly using prompts, phone numbers, and emails to target ads, sparking privacy and trust controversies.

Paragraph Summary

According to Hacker News, ChatGPT is using users' prompts, phone numbers, and email addresses to target advertising, marking a new stage in conversational AI commercialization. The article's analysis points out that prompts naturally contain high-value intent signals, and combined with strong identifiers, they can build far more precise profiles than traditional cookies, significantly increasing intrusiveness. There is a fundamental tension between users' expectations of ChatGPT's privacy and its positioning as an advertising platform, which could trigger a trust crisis. The high operating costs of AI assistants are driving ad monetization, but competitors' ad-free strategies could become a trigger for user migration. The article suggests that users clear conversation history and use separate contact information to reduce exposure, while developers need to take data collection boundaries and informed consent more seriously.