– OpenAI’s GPT Store is flooded with potentially copyright-infringing and academically dishonest GPTs
– Developers must verify their profiles and submit GPTs for review, but the moderation process appears to be lacking
– GPTs in the store include copyright infringing content, tools to bypass AI content detectors, impersonations of public figures, and attempts at jailbreaking OpenAI’s models.
OpenAI CEO Sam Altman introduced GPTs, custom chatbots powered by generative AI models, at the company’s developer conference. These bots can perform a variety of tasks, but the GPT Store, OpenAI’s marketplace for GPTs, is filled with potentially copyright-infringing bots that raise concerns about moderation.
Developers must pass OpenAI’s review process to list GPTs in the store, and building GPTs is accessible to those without coding experience. However, the store has grown rapidly, compromising quality and adherence to OpenAI’s terms. Some GPTs in the store are based on popular franchises without authorization.
Additionally, some GPTs in the store promote academic dishonesty by claiming to bypass AI content detectors. There are concerns about copyright infringement, impersonation, and jailbreaking within the GPT Store. OpenAI’s safe harbor protection shields the company from liability for copyright infringement by GPT creators.
Despite OpenAI’s efforts to create a curated marketplace for AI tools, the GPT Store is facing growing pains, including spam, legal issues, and challenges in attracting users. Monetization options for GPT developers could further complicate matters, raising questions about copyright violations and revenue sharing. The GPT Store’s shortcomings highlight the need for improved moderation and safeguards to ensure the platform’s integrity and compliance with ethical standards.