Reports: OpenAI Allegedly Paying Contractors to Read Users' Chats
The Facts
- OpenAI is allegedly paying hundreds of outside contractors more than $50 an hour to read and rate real ChatGPT conversations under an internal program called Project Lily, according to a 404 Media investigation published Monday.
- Reviewers score up to four candidate responses on a scale of 1 to 7, judging clarity and tone rather than factual accuracy. An FAQ given to the workers says separate project teams handle content verification.
- Internal documents reportedly show contractors training ChatGPT not to anthropomorphize itself and to be less sycophantic. Lawsuits allege that the over-sycophantic GPT-4o system contributed in part to multiple users' suicides.
Sources Split
LEFT
CENTER
RIGHT
+1
+1
ESTAB
CRIT
CENTER
PRO
ESTAB
The Spin
Industry-critical narrative
Nearly a billion people pour their health scares, breakups and work secrets into ChatGPT, believing it's a private conversation, and instead hundreds of outside contractors are sitting there reading the whole thread. Burying that reality in settings that default to sharing is a trap. A scrubbing filter that the company admits leaks sensitive details is no protection at all.
Pro-industry narrative
OpenAI has already spelled this out on its website: real people may review chats to improve systems and enforce safety rules, and content shouldn't include anything too sensitive to be seen. A switch to turn training off sits right there in Data Controls, alongside temporary chats that are never used for training. Human feedback is exactly how a chatbot stops being sycophantic and dangerous.
