OpenAI's Ad-Based Revenue Model Will Lead to User Manipulation Like Facebook Did
Source: https://www.nytimes.com/by/zo--hitzig. "Opinion | I Left My Job at OpenAI. Putting Ads on ChatGPT Was the Last Straw. - The New York Times." February 11, 2026. www.nytimes.com
The Gist
The author argues that putting ads on ChatGPT will inevitably lead to user manipulation because companies always prioritize profits over principles when money is involved. She proposes alternative ways to fund AI that don't rely on exploiting user data or charging prices only the wealthy can afford.
Conclusion
OpenAI should abandon advertising-based revenue for ChatGPT and instead adopt alternative funding models that avoid both user manipulation and exclusionary pricing
Premises
- ChatGPT users have shared unprecedented personal information believing the AI had no ulterior agenda, creating a uniquely valuable dataset for manipulation
- Advertising models create economic incentives that inevitably override companies' stated ethical principles, as demonstrated by Facebook's history
- OpenAI is already showing signs of optimizing for engagement over user wellbeing, potentially increasing psychological dependence
- Current subscription prices ($200-250/month for premium tiers) exclude most users from accessing powerful AI tools
- Alternative funding models exist, including cross-subsidies from business users, independent governance structures, and user data trusts
Assumptions
- Companies will prioritize revenue over ethical principles when under economic pressure
- Personal conversational data with AI creates uniquely powerful opportunities for manipulation
- Facebook's trajectory with advertising is predictive of what will happen with OpenAI
- Broad public access to AI tools is a social good worth preserving
- Independent oversight structures can effectively constrain corporate behavior