OpenAI will soon test ads in ChatGPT despite CEO Sam Altman once calling the idea dystopian

OpenAI says it will start testing ads in ChatGPT in the coming weeks. The move aims to help offset massive infrastructure costs, despite CEO Sam Altman's earlier objections to the idea.

OpenAI will introduce advertising in ChatGPT. The company plans to begin testing with logged-in adult users in the US over the coming weeks. For now, only the free tier and the new Go subscription will show ads. The pricier Pro, Business, and Enterprise plans remain ad-free.

Ads will appear at the end of ChatGPT responses when a relevant product or service matches the conversation. OpenAI says advertising will always be clearly labeled and kept separate from the actual answers, with no influence on the content itself.

Users can also find out why they're seeing a specific ad and dismiss it. Conversations with ChatGPT stay private and won't be sold to advertisers, OpenAI says. During the test phase, ads won't appear for accounts where the user indicates they're under 18 or where OpenAI predicts they might be a minor. Sensitive topics like health, mental health, and politics will also remain ad-free.

Altman reverses course on advertising

The decision marks a notable shift. In the past, CEO Sam Altman said he didn't want advertising in ChatGPT for aesthetic reasons alone. He also warned it was easy to imagine a dystopian future where ChatGPT tells you which product to buy or where to go on vacation.

OpenAI is now framing things differently: advertising will make AI more accessible. More people could benefit from the tools without paying or hitting strict usage limits. The ads are meant to help "make powerful AI accessible to everyone," as the company puts it.

OpenAI says it sees potential for new ad formats in AI-powered advertising. Conversational interfaces let users go beyond static messages and links. According to the company, users will soon be able to see an ad and directly ask questions they need to make a purchase decision.

The Information recently reported that OpenAI is already exploring whether ChatGPT's memory function could personalize ads. This feature picks up personal details from conversations, like where you live, names, or pets, and could theoretically power more targeted product recommendations.

Google is also testing "personal intelligence," which gives the AI system access to all your Google data. Google will likely use this data blending for more targeted advertising eventually. Products like Google Search and YouTube show how ads can start off harmless and then completely spiral out of control under constant quarterly revenue growth pressure.

Economic pressure likely is driving OpenAI's decision. The company needs to justify its roughly $500 billion valuation to investors, and possibly as much as $750 billion in the near future. Meeting those expectations will require unprecedented revenue growth in the coming years.

While OpenAI describes its enterprise and subscription business as "already strong," it sees advertising as part of a diversified revenue model. According to estimates from early last year, only around five percent of users currently pay for ChatGPT.

ChatGPT Go launches worldwide

Alongside the advertising announcement, OpenAI is expanding access to its cheaper ChatGPT Go subscription. The plan has been available in 171 countries since August and is now launching in the US and everywhere ChatGPT is available. For $8 per month, users get extended access to messaging, image creation, file uploads, and the memory function. They'll also see ads.