OpenAI’s GPT Store flooded with spam and copyright-infringing GPTs, raising questions over moderation | Tech News

OpenAI’s GPT Store flooded with spam and copyright-infringing GPTs, raising questions over moderation

OpenAI's GPT Store faces spam and copyright concerns with bots impersonating public figures and violating policies, despite moderation efforts.

| Updated on: Mar 21 2024, 20:30 IST
Microsoft to unveil new Copilot features! DALL-E 3 to GPT-4 Turbo, check benefits now
GPT Store
1/6 Microsoft’s take on bringing AI services to their customer has drastically shifted over the years. First, the company introduced the generative AI tool called Bing Chat. Now, Microsoft has announced an umbrella of AI features under Microsoft Copilot. The company also announced the new upcoming features to the platform for 2024. (Microsoft)
image caption
2/6 GPT-4 Turbo: Microsoft has revealed that the copilot will be integrated with OpenAI's latest model, GPT-4 Turbo. This will immensely benefit users to generate advanced responses along with managing complex and longer tasks. After the testing of the feature is complete, it will be rolled out to users in the coming weeks.  (Microsoft)
GPT Store
3/6 Advanced DALL-E 3 Model: With  DALL-E 3, the copilot will be able to generate even higher-quality images. It is said to generate accurate images based on the prompts given to the system. The feature has already been available to users on Microsoft Paint. (Microsoft)
GPT Store
4/6 Multi-Modal with Search Grounding: Microsoft is integrating the capabilities of GPT-4 to vision alongside Bing image search and web search data for better image interpretation and understanding. This will enable users to get better image queries. The feature will soon be made available to the users.  (Microsoft)
GPT Store
5/6 Deep Search: With GPT-4, Deep search will enable users to get optimized search results for complex topics. The feature broadens search queries and provides more comprehensive descriptions with relevant results. (Microsoft)
GPT Store
6/6 Additionally, there are two new features such as Inline Compose with rewrite menu which enable users to rewrite website data. The other feature is a Code Interpreter which is designed to perform complex tasks such as writing codes, data analysis, visualization, math and more.  (Microsoft)
GPT Store
icon View all Images
OpenAI's GPT Store faces spam and copyright woes, raising concerns amid plans for developer monetisation. (AFP)

OpenAI's ChatGPT Store is facing a deluge of spam and copyright concerns. The marketplace, designed for custom chatbots powered by OpenAI's generative AI models, is grappling with an influx of peculiar and potentially copyright-infringing bots, indicating lax moderation efforts.

Among the problematic bots are those claiming to generate art based on Disney and Marvel properties, serving as conduits to third-party paid services, and advertising the ability to circumvent AI plagiarism detection tools, Techcrunch reported.

Instances of bots simulating conversations with public figures without consent and attempts at "jailbreaking" OpenAI's models to increase their permissiveness have also been identified.

Also read: Sam Altman's OpenAI may bring big upgrades with GPT-5; Know how it will improve ChatGPT

OpenAI's stance

According to an OpenAI spokesperson, GPTs intended for academic dishonesty, including cheating, are against their policy. They are also against the creation of GPTs that impersonate individuals or organizations without their consent or legal rights.

However, despite these policies, the ChatGPT Store hosts numerous bots that claim to represent or imitate public figures like Elon Musk and Donald Trump without consent, raising questions about the definition of impersonation and parody.

Copyright concerns

Several bots in the ChatGPT Store seem to be extracted from popular franchises such as Star Wars, Monsters Inc., and Avatar: The Last Airbender, leading to potential copyright issues.

Despite OpenAI's terms prohibiting the promotion of academic dishonesty, the store features bots suggesting they can bypass plagiarism detectors, while others attempt to "jailbreak" OpenAI's models, though with limited success.

Originally envisioned as a curated collection of productivity tools, the ChatGPT Store has now become overrun with spam and legally dubious bots, posing significant challenges as OpenAI plans to enable developers to monetize their creations.

Also read: Google researchers leverage AI technology to forecast floods in India, potentially saving lives

OpenAI's response

An OpenAI spokesperson acknowledged the issues, mentioning they employ a combination of automated and human review processes to identify policy violations, which may result in warnings, restrictions, or removal from the store.


Catch all the Latest Tech News, Mobile News, Laptop News, Gaming news, Wearables News , How To News, also keep up with us on Whatsapp channel,Twitter, Facebook, Google News, and Instagram. For our latest videos, subscribe to our YouTube channel.

First Published Date: 21 Mar, 20:30 IST