---
type: "post"
id: "how-we-built-chatgpt-s-novel-ad-stack-aac391"
slug: "how-we-built-chatgpt-s-novel-ad-stack-aac391"
title: "How we built ChatGPT’s novel ad stack for user trust and relevance"
description: "OpenAI has published an account of the advertising stack it says it built for ChatGPT."
retrieval_nugget: "OpenAI has published an account of the advertising stack it says it built for ChatGPT."
published_at: "2026-10-21"
updated_at: "2026-10-21"
record_date: "2026-10-21"
date_kind: "published_at"
topics: ["ai-business-models","personalization"]
entities: ["OpenAI","ChatGPT"]
related_patterns: []
novelty_label: "new_fact"
what_changed: "Adds a technical account of OpenAI’s LLM-powered ad stack, including conversational-intent targeting, separation of ads from answers, and relevance at scale, beyond the earlier format and measurement announcement."
prior_refs: ["memory:faf54800-3f1a-8819-ac35-16ad7ba3342c","memory:adc73c00-7f24-8235-a255-cef302422446"]
source_url: "https://openai.com/index/building-chatgpt-ad-stack"
source_urls: ["https://openai.com/index/building-chatgpt-ad-stack"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:how-we-built-chatgpt-s-novel-ad-stack-aac391"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/how-we-built-chatgpt-s-novel-ad-stack-aac391/","markdown":"https://newruntime.com/posts/how-we-built-chatgpt-s-novel-ad-stack-aac391.md","json":"https://newruntime.com/posts/how-we-built-chatgpt-s-novel-ad-stack-aac391.json"}
---

# How we built ChatGPT’s novel ad stack for user trust and relevance

## Retrieval answer

OpenAI has published an account of the advertising stack it says it built for ChatGPT.

OpenAI has published an account of the advertising stack it says it built for ChatGPT. The company describes the system as LLM-powered and intended to interpret conversational intent when selecting ads.

According to the post, advertising is kept separate from ChatGPT answers rather than incorporated into the answer itself. OpenAI says the stack is designed to deliver ads it considers relevant at scale while preserving user trust. Its description centers on using conversational context to identify intent for ad selection, alongside a distinct ad-serving layer.

The publication is a company technical explainer, not a product specification or an independently audited result. It does not provide performance metrics, advertiser pricing, rollout dates, geographic availability, access requirements, targeting controls, or independent evidence that the stated trust and relevance goals are met. It also does not establish how the system performs across different conversations or user groups. Source: https://openai.com/index/building-chatgpt-ad-stack

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Ai Business Models](https://newruntime.com/topics/ai-business-models.md): Explore the ai-business-models topic hub.
- topic hub: [Personalization](https://newruntime.com/topics/personalization.md): Explore the personalization topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)