Ads have come to ChatGPT: what they are, how they work, and whether your business should go there
OpenAI has started showing ads right inside ChatGPT - for a business that's both a new sales channel and a new field of risks. The main new one is "conquesting": a competitor's ad can surface in a conversation about your brand. The expert consensus: organic first, ads second.
A year ago, advertising in ChatGPT seemed like something out of the future. Now it's reality: OpenAI has started showing ads right inside the chat - and is quickly building a full-fledged ad platform out of it. For a business, this is both a new sales channel and a new field of risks. Let's break it down calmly: what it is, how it works, what the "conquesting" phenomenon that marketers talk about is, and most importantly - whether your business should go there.
What it is and how it works
OpenAI officially began testing ads in ChatGPT on February 9, 2026, and has been rapidly expanding them since. It works like this: ads are matched to the topic of your conversation - so if you're discussing, say, choosing running shoes, a relevant ad may appear nearby.
OpenAI highlights several principles that are important to understand. Ads are always labeled as "sponsored" and visually separated from the answer - they aren't passed off as ChatGPT's opinion. Ads don't influence the answers themselves: according to the company, answers are optimized only for what's helpful to the user. And privacy: advertisers don't get access to your chats, history, or personal data - only anonymized performance statistics. Another important nuance: ads are seen by users on the free and entry-level tiers, while on paid ones (Pro, Business, Enterprise) there are none.
What the ad platform can already do
In just a few months, OpenAI has managed to turn this into a recognizable advertising tool, similar to Google's and Meta's dashboards. There are already custom audiences (you can upload a customer list by email or phone and target them), automatic generation of ad variations (the AI suggests creatives for approval), and flexible audience-based bids.
At the same time, the platform is still clearly "raw": matching against uploaded lists is partial, and technical glitches sometimes surface - for example, incorrect attribution, where a competitor's site is mistakenly "credited" in an answer. This is a young tool that changes literally every week.
What "conquesting" is and why it matters
Here's the point that especially worries brands. In ChatGPT, so-called conquesting is possible - when a competitor's ad can surface in a conversation about your brand. Imagine: a person asks ChatGPT about your company or product - and next to it appears an ad from your rival, sometimes even with a direct comparison not in your favor. This moves competitive rivalry into the very "conversation" with the AI and means that even your loyal customers might see someone else's offer at the most inopportune moment.
The conclusion is simple: even if you're not advertising in ChatGPT yourself yet, it's worth understanding what can be happening with your brand there.
Pitfalls worth knowing about
- The audience is skewed. Ads are seen by users on free tiers - and that, by experts' estimates, is more like students and young professionals. The most solvent users move to paid ad-free tiers. For B2B or brands aimed at an older audience, this is a risk of "missing" their public.
- The data is still approximate. Matching against uploaded audiences is only part of the list, so treat these segments as a guideline, not a precise sample.
- The platform is immature. Glitches, changing rules, the absence of established benchmarks - a normal price for being "early."
Whether your business should go there
Here experts have a surprisingly unified answer: organic first, ads second. Analysts at Monks frame it as the "duality of visibility": if your brand isn't in ChatGPT's organic answer itself, then the ad will be perceived as an intrusive interruption. So first you need to work on getting the AI to actually know and "recommend" you (this is called AEO/GEO - optimization for AI answers), and only then bring in paid tests.
In practice, this means the following. Assess whether the audience is yours - if you're in B2B or work with a mature, solvent public, maybe there's no rush. Prepare your content so the AI can "read" and recommend you: clear answers to common questions, transparent prices, honest comparisons with competitors. And if you try ads - go in as a test with a small budget and different success metrics: here, what matters is not the number of clicks but the match with the person's intent.
The key points in brief
OpenAI has launched ads in ChatGPT: they're matched to the topic of the conversation, labeled as sponsored, don't influence answers, and are shown only on free tiers. The dashboard already resembles Google and Meta - with custom audiences and creative auto-generation - but the platform is still raw. The main new risk is conquesting, when a competitor surfaces in a conversation about your brand. Should you go there? First - organic visibility (so the AI knows and recommends you), and only then - careful paid tests, keeping an eye on whether the audience there is yours. Simply put: you already need to watch this space, but invest deliberately and without rushing.
Advertising is increasingly moving to where people ask AI their questions - and the rules of the game there are still taking shape. To figure this out ahead of competitors and avoid wasting budget - subscribe to my newsletter. I explain important updates from Meta, Google, OpenAI, and other platforms in plain language and always with a concrete takeaway: what exactly you should do about it.
Subscribe to the newsletter