ChatGpt's digital store is flooded with spam and software in violation of copyright and the rules of the platform itself, some leading to paid services.
This was discovered by the specialized technology site TechCrunch, which raises doubts about the control and moderation capabilities of the parent company, OpenAI.
The Gpt Store launched in November, building on the success of the AI chatbot, and is the online showcase for ChatGpt-powered digital assistants.
The portal works as a sort of app store for all individuals and companies who want to offer services based on interaction with the now well-known software.
Creating a chatbot requires no programming experience, the site notes, and developers can type the features they want to offer into the tool.
Perhaps due to the low barrier to entry, the GPT Store has grown rapidly and now has 3 million GPT software, a growth that appears to have come at the expense of quality, however, as well as compliance with OpenAI's own terms, says TechCrunch.
According to the site, in fact, on the digital store there are chatbots that can create images of Disney and Marvel characters, in violation of copyright.
Or software that simulates conversations with famous people including Elon Musk, Donald Trump and Leonardo DiCaprio, and impersonation is prohibited by OpenAI rules.
Still other software leads to paid third-party services and advertises itself as being able to bypass AI content detection tools like Turnitin and Copyleaks.
In its moderation rules, OpenAI explains that it uses a combination of automated systems, human review and user reporting to spot chatbots that violate the rules.
Violations may lead to actions against content or an account such as warnings, restrictions, or ineligibility for inclusion on the Gpt Store.
Reproduction reserved © Copyright ANSA