A recent episode of the 404 Media podcast highlights a little-known but significant practice: human contractors are reading actual ChatGPT conversations. According to the show, these workers review real user prompts, which means that anything typed into the chatbot could be seen by human eyes. This raises obvious privacy questions, especially for users who share sensitive or personal information in their chats.
The podcast also touches on two other tech industry stories. First, it covers Automattic, the company behind WordPress.com, which apparently made a major policy change and then reversed it after public pressure — a pattern the hosts describe as an 'out-and-back-in.' Second, it discusses a16z, the venture capital firm, which claims that the concept of 'enshittification' — the gradual degradation of online platforms — is not real. That stance directly contradicts the experience of many users and critics who see quality decline as a common pattern.
Because this is a single source, we can only report what the podcast says. The three topics are distinct, but together they paint a picture of an industry where user trust is often secondary to business decisions, and where even the most popular AI tools are not as private as they might seem.