Former OpenAI researcher Zoe Gitzig said that the introduction of advertising in ChatGPT could turn the service into a product with the same incentives that once pushed Facebook to make vague privacy promises.
Former OpenAI researcher Zoe Gitzig said that the introduction of advertising in ChatGPT could turn the service into a product with the same incentives that once pushed Facebook to make vague privacy promises.
According to Ars Technica, Gitzig filed for resignation on Monday, the day OpenAI began testing ad units in ChatGPT in the U.S. She wrote about her decision in a guest column for The New York Times, emphasizing that advertising itself is not immoral, but the danger lies in the data the targeting system can obtain in the conversational product.
She believes ChatGPT has amassed an unprecedented archive of human candor over the years: users share medical fears, relationship issues, and religious beliefs because they see the chatbot as a tool with no hidden agenda. If you build an advertising engine around that, the pressure to optimize for clicks and retention rather than usefulness is there, and with each iteration, the rules can become more relaxed.
«This week has confirmed my slow realization: OpenAI seems to have stopped asking the questions I came here to ask,» said Gitzig.
The report also says that OpenAI planned to display ads at the bottom of responses, label them, and prevent ads from influencing the model’s responses. The test applies to the free tier and Go subscription, while Plus, Pro, Business, Enterprise, and Education plans will remain ad-free. However, the company’s documentation describes ads personalization based on current and previous chats, unless the setting is turned off.
Hitzig suggests alternatives to the advertising model, including cross-subsidizing access, independent oversight with mandatory decisions on the use of voice data, and mechanisms where control over information remains with users. She believes the industry must choose between a free product that manipulates and a service that is only available to those who can pay.
Previously, dev.ua reported that OpenAI fired its vice president of product policy, Ryan Beyermeister, in January. She was responsible for ChatGPT’s terms of use and other product policies, including moderation approaches. The company explained the decision by citing an internal complaint from a male colleague about alleged gender discrimination.
