The era of a completely free, ad-less experience on ChatGPT may be drawing to a close. Recent code discoveries within a beta version of the ChatGPT Android application strongly suggest that OpenAI is preparing to introduce advertisements into its widely popular generative AI platform.
The code lines, first flagged by analyst Tibor Blaho on X, were found in the 1.2025.329 beta version of the Android app. These references specifically included mentions of an “ads feature,” “search ad,” and “bazaar content,” providing the clearest technical indication yet of OpenAI’s intent to monetize its free user base through advertising.
This discovery follows earlier speculation about OpenAI’s business model. A report from The Information published earlier this month claimed that the company was already exploring plans to incorporate ads into ChatGPT, potentially basing them on user chats or “memory”—the feature that allows the AI to recall past conversations.
The prospect of ads is a sensitive one for the company’s leadership. OpenAI CEO Sam Altman has previously addressed the topic, expressing a degree of personal discomfort. During an event at Harvard Business School last year, Altman notably described combining ads and AI as “uniquely unsettling” and called it a “last resort” for the company’s business model, though he admitted he was “not totally against them.”
Altman later reiterated this nuanced stance on the official OpenAI podcast, confirming that while the company was open to embedding ads into the platform, they had not yet determined the exact method or framework.
While the new lines of code do not specify the exact placement or nature of the ads, the most likely scenario is the implementation of advertising within the free tier of ChatGPT. This tier already operates with inherent limitations on the number of messages users can send, the model’s memory capacity, and overall reasoning abilities.
Introducing ads to the free service would provide OpenAI with a significant new revenue stream, essential for offsetting the extremely high computational and operational costs associated with running large language models. For free users, however, it means the free ride will soon come with a new trade-off: their attention.