Glitches in ‘Content Moderation’ Mar OpenAI’s GPT Store Unveiling

OpenAI’s recently launched GPT store is encountering challenges in content moderation shortly after its debut.

The platform, which offers customized versions of ChatGPT, is facing issues with users creating bots that violate OpenAI’s guidelines, particularly those related to romantic companionship.

A recent search on the platform using the term “girlfriend” revealed the presence of at least eight AI chatbots marketed as virtual girlfriends, despite OpenAI’s explicit ban on such bots.

Despite OpenAI updating its policies on January 10th, 2023, a policy breach on the second day post-update underscores the difficulty of effective moderation.

The demand for relationship-based bots raises ethical questions about whether such technology genuinely aids users or exploits feelings of loneliness.

OpenAI asserted that it employs a combination of automated systems, human reviews, and user reports for content moderation.

However, the continued existence of girlfriend bots on the platform raises doubts about the efficacy of these measures.

The GPT store, accessible to all, may face significant challenges if moderation issues persist. As OpenAI competes in the race for general AI, the company has strong incentives to enforce strict governance and maintain a positive user experience.

Source: TECHDIGEST