Former OpenAI researcher Zoë Hitzig warned that strong commercial incentives could eventually erode early safeguards, drawing parallels with the evolution of advertising practices at Facebook.
OpenAI says advertising in ChatGPT is intended to help fund free access to advanced artificial intelligence tools, but early reactions from researchers and commentators suggest the strategy could test the company’s ability to balance revenue with user trust, according to analysis published by WARC.
In interviews and public statements, OpenAI executives have framed advertising as a way to “democratize access” to AI by subsidizing high usage limits for free users. Ads are currently being tested only for users on free and entry-level plans, while paid subscribers and enterprise customers are not shown advertising.
The company has said the advertising system operates separately from the AI model itself and does not influence responses. Internal decision-making prioritizes user trust above user value, advertiser value and revenue, according to OpenAI’s monetization leadership, and ads judged intrusive or overly personalized are intended to be rejected.
Personalization is also limited in sensitive areas such as health or politics, and users are able to review or delete the data used for ad targeting or turn off personalization altogether.
Even so, the long-term durability of those principles is being questioned. Writing in The New York Times, former OpenAI researcher Zoë Hitzig warned that strong commercial incentives could eventually erode early safeguards, drawing parallels with the evolution of advertising practices at Facebook.
Hitzig argued that while companies often begin with strict rules to protect users, the pressure to scale revenue can gradually reshape those boundaries as platforms grow.
Industry analysts say the economic rationale for advertising in AI services is clear. Training and operating large language models requires significant investment, and consumer research cited by WARC suggests roughly four in five users are willing to accept advertising in exchange for free access to AI tools.
The comparison with Facebook is frequently raised in industry discussions. Despite repeated controversies over privacy and content moderation, the social media platform has built one of the world’s most profitable advertising businesses, supported by billions of daily users.
Whether conversational AI platforms can replicate that model remains uncertain. Unlike social networks, chatbots operate in a more intimate, task-focused environment, where poorly targeted or intrusive advertising could undermine user confidence more quickly.
For now, OpenAI says it is taking a cautious approach, showing only a limited number of ads while testing formats and user reactions. Analysts say the coming years will reveal whether the company can maintain its stated priorities—or whether, as with earlier internet platforms, the demands of scale begin to reshape the balance between trust and monetization.
