OpenAI's forced pivot to advertising reveals a fundamental flaw in the business model of LLMs — the technology simply isn't economically viable without ads. The shift transforms ChatGPT from a tool serving users into a media company selling attention to advertisers, introducing perverse incentives where AI outputs will be manipulated to favor paid brands rather than providing neutral, factual responses to users. ChatGPT's memory feature amplifies these concerns by potentially enabling the exploitation of sensitive personal data.
The compute deals enabling OpenAI's expansion aren't risky — what's risky would be failing to secure the infrastructure needed to deliver transformative products that could give everyone access to services previously available only to the wealthy. Advertising represents just one potential revenue stream among massive market opportunities in consumer and enterprise AI, and any implementation would require a fundamentally different, privacy-respecting model than traditional ad platforms.
© 2025 Improve the News Foundation.
All rights reserved.
Version 6.18.1