Skip to content
InYourGeek
visiteur@inyourgeek — shell
↹ compléter↑↓ historique⏎ ouvrirhelp
FR

The Monday letter · the tech and security week in one email

Free, every Monday, via Brevo (EU), one-click unsubscribe · your data and rights

AI· 2 min read

ChatGPT's ads in the US can target you by your prompts, email and phone

In the United States, ChatGPT ads can be aimed at the conversation you're having, but also at your email address or phone number. A question you put to an assistant isn't necessarily meant as an ad-targeting signal.

Illustration of a chat window with a speech bubble on one side and, on the other, a phone and an envelope with a bullseye target, suggesting ads aimed at users.

In the United States, ChatGPT’s ads don’t rely only on what you type. According to Next, an article published on 16 October 2026, an advertiser can reach you based on the conversation you’re having, but also from your email address or phone number.

Ads that read the conversation

OpenAI has been running ads inside its chatbot since the start of the year, first in the United States, then in France and Europe from the end of August. Conversation-based targeting, however, is not yet active in France.

Advertisers don’t pick a specific question. They describe a context. 404 Media tested the tool by creating two ads, and the form asks you to “Describe the conversations, topics or keywords where your products or services could be relevant”. OpenAI explains that “Contextual hints help our systems determine when your ads may be relevant in ChatGPT”. The company also encourages advertisers to run several ad types: a plumber, for instance, can separate emergency repairs from routine maintenance.

Lists of emails and numbers

The second part concerns “custom audiences”. OpenAI describes them like this: “custom audiences are created from lists of customers or prospects using email addresses, phone numbers, hashed email addresses, hashed phone numbers, or Google advertising IDs (GAID)”. So, according to Next, advertisers can use lists of phone numbers or emails without even hashing them. Asked about this by 404 Media, the company did not respond. It only said the files are deleted after processing, normally within 24 hours.

A 25,000-user threshold

These lists can act as blocklists, which exclude the people targeted, or as allowlists, which include them. The latter can only be used if they contain at least 25,000 ChatGPT users. OpenAI says that “Being on an advertiser’s list alone is not enough to trigger an ad.” According to the company, advertisers cannot see which people match.

A promise and a mechanism

OpenAI assures, according to Next, that it does not share information about its users with advertisers. The nuance lies in the direction of the flow: advertisers bring their email and phone lists, and the conversation in progress helps decide when their ad appears.

Next time you ask ChatGPT how to descale a kettle, bear in mind that in the United States, that question may also be used to choose which ad gets shown to you.

Sources (1)

Written with AI assistance from the sources cited above, then reviewed and approved before publication by Sébastien Soulier.

The Monday letter

The tech and security week, every Monday in your inbox

What to patch first, data breaches, the articles with pictures. Free, no ads, one-click unsubscribe.

Free, every Monday, via Brevo (EU), one-click unsubscribe · your data and rights