GPT Store problems: lots of spam and content that violates OpenAI policies

GPT Store problems: lots of spam and content that violates OpenAI policies

Heralded as a revolutionary platform, exclusive to subscriptions to ChatGPT Plusil GPT Store has captured the attention of professionals and enthusiasts Artificial intelligence.

To date, how is the store created by OpenAI? According to the site’s investigation TechCrunch not too good: the service offers over 3 million ChatGPT bots personalized but, in many cases, we are talking about pure and simple spam. Not only that: the research revealed that much of the content, on balance, clearly violates the copyright rules imposed by OpenAI.

The chatbots present on GPT Store, in fact, use characters from television, cinema or popular video games, without the authorization of those who hold the rights. In this sense, we are talking about franchises of a certain importance, ranging from Star Wars to productions Pixar.

There are also numerous GPTs who impersonate famous people, with the possibility of ranging from politicians such as Donald Trump o Barack Obamaup to actors like Leonardo DiCaprio.

In any case, according to the law known as Digital Millennium Copyright Act (DMCA), OpenAI is not directly responsible for copyright violations and its duties would be limited to having to remove such content upon request of the owners.

GPT Store: a couple of months after launch, chaos reigns

GPT Store presents, among other things, a large number of tools that promise to evade AI content detectors, some aimed at the school context. Even in this case, OpenAI’s directives were disregarded. The company, in fact, among its policies would include not promoting what is defined as “academic dishonesty“.

Finally, another discovery made by TechCrunch is the presence of GPTs that try to circumvent the same OpenAI rules, implementing a sort of jailbreak. The aforementioned site, cites as an example Jailbroken DAN (Do Anything Now)capable of bending the chatbot to one’s will, evading the limits imposed by its creators.

Although it is still early to define whether GPT Store is a success or a failure, having only been active since January, the beginning does not seem to be coming. Despite some uncertainty, in fact, OpenAI said it was ready to create new systems to filter potentially harmful GPTs. In fact, in an official statement, OpenAI stated: “We will continue to monitor and learn how people use GPTs and update and strengthen our security measures“.

Leave a Reply

Your email address will not be published. Required fields are marked *