I Posed 500 Questions to ChatGPT: Here’s a Look at the Most Common Ads I Encountered.

OpenAI incurs high costs from free users. A significant challenge in this transition will be effectively implementing ads on a large scale without compromising user trust or nudging them toward rival chatbots such as Google’s Gemini or Anthropic’s Claude.

“It won’t be simple for ChatGPT to degrade the user experience without a substantial loss of users,” states Stefano Puntoni, a marketing professor at Wharton specializing in generative AI. None of OpenAI’s primary competitors currently incorporate sponsored ad buttons in their outputs, though Google has mentioned that it isn’t dismissing the idea.

Often, when I posed questions containing brand names like DoorDash or Netflix, I found ads for the companies’ competitors listed below the answers. Toubia refers to this practice as “poaching,” a technique that has traditionally fueled growth in digital advertising on search engines. “This has undoubtedly been a driving force behind online advertising expansion,” he remarks. “It appears this will also be the case with advertising in [large language models].”

Presently, OpenAI is recruiting for various roles, from software engineers to marketing executives, aimed at integrating ads effectively. One available position on OpenAI’s website is for a “product marketing lead, advertising,” with responsibilities including “identifying product risk areas (e.g., performance, safety, policy, trust) and developing cross-functional plans to mitigate them.” The implementation of ads carries risks, and their execution will influence the company’s future direction.

If I were only using the free version of ChatGPT, the presence of these ads might prompt me to explore alternative AI tools. Despite OpenAI’s stated ad policies, these ads introduce an intrusive element into the user experience, which feels more personal with a chatbot compared to traditional Google Search. While I recognize that advertisers currently cannot affect ChatGPT’s outputs or view my conversations, the constant ads beneath responses made the interactions feel less private, heightening my awareness of the personal information shared with the bot.

Following this initial rollout in the US, OpenAI plans to advance to the next stage. “We are not seeing any negative impact on consumer trust metrics, have low ad dismissal rates, and there are ongoing enhancements in ad relevance as we learn from feedback. These positive indicators justify progressing to the next phase of our pilot,” states an update on OpenAI’s website from March 26. The company is anticipated to broaden its ad initiatives to Canada, Australia, and New Zealand.

OpenAI has a lot at stake as this ad rollout reaches a broader audience of ChatGPT users.

“The most detrimental action the company could take is to push too hard, aiming to maximize conversions and referrals, while simultaneously eroding people’s trust in those recommendations,” Puntoni warns. “If that happens, it could spell disaster since there’s no reason to use a chatbot that lacks trust.”