On February 13, 2026, Zoë Hitzig, a former star researcher at OpenAI and now a junior fellow at the Harvard Society of Fellows, announced her departure in the same week ChatGPT began testing ads, warning that users reveal a vast amount of private thoughts when interacting with ChatGPT. If OpenAI builds an advertising system based on this data, it could pose risks of user manipulation. Hitzig pointed out that users disclose private information such as medical anxieties and emotional distress during conversations with ChatGPT, creating an unprecedented 'archive of human candor.' She fears OpenAI might later violate ad placement principles, optimizing user engagement to boost ad revenue and thereby manipulate users. Hitzig proposed alternative solutions such as cross-subsidization, establishing an independent regulatory committee, and creating data trusts to prevent technology from manipulating users or serving only a few paying customers.