OpenAI Begins US Test of Visual Ads Next to ChatGPT Image Results

OpenAI Announces Visual‑Ad Pilot

On October 5, 2026 OpenAI disclosed a new advertising product that places visual ads next to the images produced by its generative AI model within ChatGPT. The company described the format as a “side‑by‑side” placement that will be clearly labelled and physically separated from the AI‑generated content.

The rollout is scheduled to begin later in October and will be limited to a small group of advertisers operating in the United States. OpenAI has not announced any timeline for extending the test beyond U.S. borders, and no European launch has been confirmed.

Scope and Audience

According to the announcement, the ad format will only be shown to users who are 18 years of age or older and who are using either the free tier or the low‑price subscription plans of ChatGPT. Users on higher‑price enterprise plans will not see these ads, and the company says the placements will be invisible to minors.

OpenAI also emphasized that the ads will not influence the image generation process. The AI model will continue to produce results based solely on the user’s prompt, and the ad layer will be added afterwards, ensuring a strict separation between content and commercial messages.

Safety and Brand‑Safety Measures

To avoid inappropriate placements, OpenAI said it will employ a combination of automated filters and human review. The filters are designed to block ads from appearing in contexts deemed sensitive, such as medical, political or extremist topics.

The company also named three third‑party partners that will help verify the safety and suitability of each ad slot. AppsFlyer will provide measurement data, while DoubleVerify and Integral Ad Science will handle brand‑safety verification and fraud detection.

Potential Reach

OpenAI claims that ChatGPT reaches roughly 1.2 billion users on a weekly basis. This figure comes directly from the company and has not be independently audited, but it gives a sense of the scale of the audience that could eventually be exposed to visual ads if the test proves successful.

• Limited to U.S. market during the pilot phase
• Targeted at adult users of free or low‑cost plans
• Ads clearly labelled and visually separated from AI output
• Human and automated review to block placements in sensitive contexts

OpenAI’s decision to start with a visual ad format marks a shift from its earlier reliance on subscription revenue and API licensing. The company has long positioned itself as a developer‑first platform, but the growing competition from other AI providers has prompted a search for additional monetisation channels.

Industry observers note that the test could provide valuable data on user tolerance for advertising within generative AI workflows. If users accept the format without a noticeable drop in engagement, advertisers may see a new premium inventory that combines high intent (users actively creating visual content) with a captive audience.

For English‑speaking organisations, the practical implication is clear: a successful rollout could open a channel to reach millions of professionals, marketers and hobbyists who rely on ChatGPT for visual brainstorming. Brands will need to craft concise, visually appealing creatives that respect the platform’s strict labeling rules, while also preparing for the additional compliance checks required by DoubleVerify and Integral Ad Science.

The pilot will be restricted to adult users who are on the free tier or the low‑cost subscription, meaning that the majority of premium enterprise customers will remain ad‑free. This segmentation aligns with OpenAI’s stated goal of keeping the higher‑value enterprise experience untouched by commercial interruptions.

OpenAI has pledged transparency around the ad experience, promising to publish metrics on impression volume, click‑through rates and any reported safety incidents once the test concludes. The data will be shared with advertisers and the broader research community to assess the impact on user behavior.

Critics have raised concerns that placing ads alongside AI‑generated imagery could blur the line between organic content and paid promotion, especially in creative domains where users may assume every visual output is purely machine‑generated. OpenAI’s labeling requirement is intended to mitigate that risk, but the effectiveness of the visual cue remains to be validated.

From a regulatory perspective, the U.S. test sidesteps the stricter advertising disclosures mandated in the European Union under the Digital Services Act. Should the format expand internationally, OpenAI will likely need to adapt its labeling and consent mechanisms to satisfy EU‑wide standards.

In summary, the visual‑ad pilot represents OpenAI’s first foray into on‑platform advertising for its consumer‑facing chatbot, targeting a massive weekly user base while attempting to preserve safety, brand integrity and a clear separation between AI output and commercial messaging.