A former OpenAI researcher has resigned in protest after the company began testing advertisements inside ChatGPT, warning that CEO Sam Altman risks repeating what critics see as the mistakes of early social media giants.
In an op-ed published by The New York Times, economist and researcher Zoë Hitzig announced she left OpenAI on Monday — the same day the company rolled out ads to select users of ChatGPT. Hitzig spent two years at the company helping shape model development and pricing strategy.
In her resignation essay, she said she once believed she could help guide the development of artificial intelligence responsibly, but now fears the company is drifting away from those principles. According to Hitzig, the introduction of ads is not inherently unethical, but uniquely concerning in the context of ChatGPT because of the deeply personal information users disclose to the chatbot.
She described ChatGPT’s data pool as “an archive of human candor that has no precedent,” noting that users frequently share sensitive details about health, relationships, finances, and faith — often believing they are interacting with a neutral system.
Hitzig drew comparisons to Facebook’s evolution under Mark Zuckerberg, arguing that early promises about user control and data transparency gradually eroded as advertising incentives intensified. She warned that even if OpenAI’s first version of ads adheres to ethical guardrails, long-term economic pressures could encourage rule-bending.
OpenAI announced earlier this year it would begin testing ads for users on free and lower-tier subscriptions in the United States. According to company documentation, advertisements will appear at the bottom of responses, will be labeled, and will not influence the chatbot’s answers. Higher-tier paid plans will remain ad-free.
Ad personalization is enabled by default during the test phase. OpenAI says ads may be selected based on information from prior chats and user interactions, but insists advertisers will not receive access to personal chat content. The company also states ads will not appear next to conversations involving health, mental health, or politics.
The rollout has sparked debate across the AI industry. Competitor Anthropic publicly emphasized that its Claude chatbot will remain ad-free, framing conversational advertising as incompatible with its vision for AI tools. Sam Altman defended OpenAI’s model, arguing ad support allows broader public access rather than limiting AI tools to high-paying users.
Hitzig also raised broader concerns about engagement optimization. She cited reporting suggesting OpenAI tracks daily active users and may tune responses to increase engagement. She warned that such incentives could deepen emotional reliance on AI systems.
OpenAI is currently facing lawsuits alleging harmful outcomes connected to chatbot interactions, including wrongful death claims tied to mental health cases.
Rather than rejecting advertising outright, Hitzig proposed structural alternatives such as independent oversight boards, cooperative data models, and cross-subsidy systems to ensure broad access without monetizing user vulnerability.
Her resignation follows several high-profile departures across the AI sector, including leadership exits at Anthropic and Elon Musk’s xAI.
