OpenAI’s CEO, Sam Altman, unveiled GPTs, bespoke chatbots powered by OpenAI’s generative AI models, at the firm’s inaugural developer conference in November. He touted them as versatile tools capable of a wide range of tasks, from coding to learning about obscure scientific topics to providing fitness advice.
Altman emphasized that GPTs, with their blend of instructions, expanded knowledge, and actions, could be more beneficial. He stated, “You can construct a GPT for virtually anything.” His statement about “anything” was not an exaggeration.
You can construct a GPT for virtually anything.” – Sam Altman, CEO of OpenAI
Investigation revealed that OpenAI’s official GPT marketplace, the GPT Store, is inundated with peculiar, potentially copyright-infringing GPTs that suggest OpenAI’s moderation efforts are lax. A quick search reveals GPTs claiming to create art in the style of Disney and Marvel properties, but they merely redirect to third-party paid services and advertise their ability to evade AI content detection tools like Turnitin and Copyleaks.
Lack of Moderation
To feature GPTs in the GPT Store, developers must validate their user profiles and submit GPTs to OpenAI’s review system, which employs a combination of human and automated review. A spokesperson explained the process: he said “our approach to identifying and evaluating potentially policy-violating GPTs involves a mix of automated systems, human review, and user reports. Violations can result in actions against the content or your account, such as warnings, sharing restrictions, or disqualification from inclusion in the GPT Store or monetization.”
A spokesperson explained the process: he said “our approach to identifying and evaluating potentially policy-violating GPTs involves a mix of automated systems, human review, and user reports.
Creating GPTs doesn’t necessitate coding skills, and GPTs can be as straightforward or intricate as the creator desires. Developers can input the capabilities they want to provide into OpenAI’s GPT-building tool, GPT Builder, and the tool will strive to create a GPT to execute those tasks.
Perhaps due to the low entry barrier, the GPT Store has expanded rapidly. OpenAI stated in January that it housed approximately 3 million GPTs. However, this growth seems to have come at the cost of quality and compliance with OpenAI’s own terms.
Copyright Concerns
The GPT Store contains several GPTs derived from popular film, TV, and video game franchises — GPTs not created or authorized by the owners of those franchises. One GPT generates monsters in the style of Pixar’s “Monsters, Inc.,” while another offers text-based adventures in the “Star Wars” universe.
These GPTs, along with others in the GPT Store that allow users to interact with trademarked characters like Wario and Aang from “Avatar: The Last Airbender,” are setting the stage for copyright disputes.
Kit Walsh, a senior staff attorney at the Electronic Frontier Foundation, clarified: [These GPTs] can be used to create transformative works as well as for infringement [where transformative works refer to a type of fair use shielded from copyright claims.] The individuals engaging in infringement, of course, could be liable, and the creator of an otherwise lawful tool can essentially talk themselves into liability if they encourage users to use the tool in infringing ways. There are also trademark issues with using a trademarked name to identify goods or services where there is a risk of users being confused about whether it is endorsed or operated by the trademark owner.
OpenAI itself would be exempt from copyright infringement liability by GPT creators thanks to the safe harbor provision in the Digital Millennium Copyright Act, which protects it and other platforms (e.g. YouTube, Facebook) that host infringing content as long as those platforms meet the statutory requirements and remove specific examples of infringement when requested. However, it’s not a good look for a company currently embroiled in litigation.
Academic Dishonesty
OpenAI’s terms explicitly forbid developers from creating GPTs that promote academic dishonesty. Yet, the GPT Store is filled with GPTs suggesting they can bypass AI content detectors, including detectors sold to educators through plagiarism scanning platforms.
Yet, the GPT Store is filled with GPTs suggesting they can bypass AI content detectors, including detectors sold to educators through plagiarism scanning platforms.
One GPT purports to be a “sophisticated” rephrasing tool “undetectable” by popular AI content detectors like Originality.ai and Copyleaks. Another, Humanizer Pro — ranked No. 2 in the Writing category on the GPT Store — claims that it “humanizes” content to bypass AI detectors, preserving a text’s “meaning and quality” while delivering a “100% human” score.
Some of these GPTs are thinly disguised conduits to premium services. For instance, Humanizer invites users to try a “premium plan” to “use [the] most advanced algorithm,” which sends text entered into the GPT to a plug-in from a third-party site, GPTInf. Subscriptions to GPTInf cost $12 per month for 10,000 words per month or $8 per month on an annual plan — a bit steep on top of OpenAI’s $20-per-month ChatGPT Plus.
We’ve written before about how AI content detectors are largely ineffective. Beyond our own tests, several academic studies show that they’re neither accurate nor reliable. However, it remains the case that OpenAI is allowing tools on the GPT Store that promote academically dishonest behavior — even if the behavior doesn’t achieve the intended result.
The OpenAI spokesperson also stated that GPTs that are designed or directed to bypass OpenAI’s protective measures or violate OpenAI’s guidelines are not permitted. However, GPTs that aim to modify model behavior in other ways — including attempts to make GPT more lenient without infringing our usage policies — are acceptable.
Developmental Challenges
OpenAI introduced the GPT Store as a carefully selected assortment of potent AI tools aimed at enhancing productivity. Despite the shortcomings of these tools, the platform is rapidly becoming a hub for spammy, legally questionable, and potentially harmful GPTs, or at least GPTs that blatantly disregard its regulations.
Given the current state of the GPT Store, the prospect of monetization could lead to a whole new set of complications. OpenAI has committed to allowing GPT developers to “profit based on the number of people utilizing “their” GPTs” and possibly even provide subscriptions to individual GPTs. But what will be the response of Disney or the Tolkien Estate when unauthorized Marvel- or Lord of the Rings-themed GPT creators start earning significant income?
OpenAI’s intention with the GPT Store is evident. As can be seen in the Apple App Store model which has been incredibly profitable, and OpenAI is essentially attempting to replicate it. GPTs are hosted, developed, promoted, and evaluated on OpenAI platforms. Moreover, they can now be accessed directly from the ChatGPT interface by ChatGPT Plus users, providing an additional reason to subscribe.
However, the GPT Store is encountering the same initial hurdles that many large-scale digital marketplaces for apps, products, and services faced. In addition to spam, a recent article in The Information highlighted that GPT Store developers are finding it difficult to attract users, partly due to the GPT Store’s inadequate back-end analytics and less than optimal onboarding experience.
The GPT Store is in disarray — and unless something changes soon, it might remain that way.
One might have expected OpenAI, given its emphasis on curation and the importance of safeguards, to have taken measures to avoid these apparent pitfalls. But that doesn’t seem to be the case. The GPT Store is in disarray — and unless something changes soon, it might remain that way.
The situation with the GPT Store is indeed concerning. OpenAI’s vision of creating a marketplace for AI tools is commendable, but the execution seems to be lacking. The issues of spam, copyright infringement, and promotion of academic dishonesty are serious and need to be addressed promptly. It’s clear that the current moderation efforts are not sufficient. OpenAI needs to invest more resources into ensuring that the GPT Store is a safe and reliable platform for both developers and users.
The promotion of academic dishonesty is a serious violation of OpenAI’s terms. It’s disheartening to see GPTs that promote such behavior, and it’s crucial that OpenAI takes swift action to remove these GPTs from the platform.
While the issues with the GPT Store are troubling, it’s important to remember that this is a relatively new platform. Many digital marketplaces have faced similar challenges in their early stages. It’s crucial for OpenAI to learn from these experiences and implement effective solutions. The potential of the GPT Store is enormous, and with the right adjustments, it could become a valuable resource for AI tools.
The issue of spam in the GPT Store is a clear indication that OpenAI needs to ramp up its moderation efforts. It’s not enough to simply have a review system in place; it needs to be effective and robust enough to catch these violations before they become a problem.
Despite these issues, I believe that the GPT Store has great potential. With the right adjustments, it could become a valuable resource for AI tools. However, OpenAI needs to act quickly to address these problems and regain the trust of its users.
The copyright issues are also a major concern. It’s one thing to create a GPT that generates content in the style of a popular franchise, but it’s another thing entirely to profit from it without the consent of the copyright holder. This could lead to legal complications for both the developers and OpenAI.
This situation is reminiscent of the early days of the App Store and Google Play Store, where spam and low-quality apps were rampant. It took a while for those platforms to implement effective moderation and quality control measures. OpenAI seems to be facing a similar challenge with the GPT Store
The copyright concerns echo the issues YouTube faced with unauthorized uploads of copyrighted content. It’s a complex issue that requires a delicate balance between protecting intellectual property rights and fostering creativity and innovation.
The situation with the GPT Store is somewhat reminiscent of YouTube’s ongoing struggle with unauthorized uploads of copyrighted content. Despite YouTube’s Content ID system and other moderation efforts, copyright infringement remains a significant issue. It’s a complex problem that requires continuous refinement of detection and enforcement mechanisms. It seems OpenAI is facing a similar challenge in moderating the GPT Store. The question is whether they will take this issue seriously and invest in more robust moderation strategies.
The GPT Store’s current predicament is a classic case of the “tragedy of the commons.” When a resource is open to all, it tends to be exploited and depleted unless there are effective rules and enforcement. This is a common issue in digital platforms. Even the reinforcement is like wrack a mole case…