OpenAI researcher quits over ChatGPT ads, warns of ‘manipulation’ risk

New Delhi: A former OpenAI researcher has quit the company and warned that introducing advertising inside ChatGPT could create new risks of user manipulation, as OpenAI begins testing ads in the US on its free and low-cost tiers.

Zoe Hitzig, an economist and junior fellow at the Harvard Society of Fellows, announced her resignation in a guest essay published by The New York Times on February 11, according to reports. Hitzig said she joined OpenAI in 2024 and worked on model development, pricing and early safety policy.

Hitzig timed her exit to coincide with OpenAI’s ad tests, which started on February 9. She argued that ChatGPT holds an unusually sensitive archive of user conversations and that an advertising model could create incentives to optimise for engagement and revenue, rather than user safety.

OpenAI said it is introducing ads in ChatGPT for some US users on the Free and Go tiers. The company said ads will appear in clearly labelled boxes below the assistant’s answer, and that the ads will not influence responses.

In its post announcing the tests, OpenAI said “answer independence” and conversation privacy will remain central, and users will have controls such as the option to turn off personalisation. OpenAI also said some paid tiers, including Plus, Pro, Business, Enterprise and Education, will remain ad-free.

The resignation adds to a wider debate across AI companies over how to balance access, monetisation and safety, as consumer AI tools become more embedded in daily decision-making and personal life.