The GPT4free project creator, Xtekky, a European computer science student, has received a letter from OpenAI demanding that he remove the entire project within five days or face legal action.
In an interview, Xtekky expressed his belief that OpenAI should not target him, as he is not directly connecting to the company’s API but rather obtaining data from other sites that have their own API licenses. He suggested that if the owners of those sites have concerns with his scripts querying them, they should address him directly.
According to Xtekky, the same result could be achieved by opening tabs of websites such as Phind or You.com and making multiple requests. “My repository just simplifies the process,” he explained.
Although GPT4Free sources its data from other sites, all of them are paying OpenAI fees to use its large language models. Thus, when using the scripts, those sites end up paying for the queries on behalf of the users without the users ever actually visiting them. If those sites rely on ad revenue to cancel these API costs, they are losing money due to these queries.
According to Xtekky, he is willing to remove any scripts that use individual sites’ APIs upon the request of the site owners. In fact, he has already taken down scripts for phind.com, ora.sh, and writesonic.com.
Additionally, Xtekky highlighted that these sites could prevent external use of their internal APIs by implementing common security measures. One potential measure You.com could take is blocking API traffic from all IPs except their own.
Xtekky claims to have advised all the sites that contacted him to secure their APIs, but none of them have taken action. Therefore, any other developer could replicate his approach even if he removes the scripts from his repository.
The developer argues that he should not be held responsible for the actions of others who use his script and that OpenAI should not target him for utilizing other sites’ unsecured APIs available on the open web.
According to Xtekky, OpenAI could have collaborated with the affected sites to issue a takedown request instead of targeting him with a legal threat. He expressed uncertainty about OpenAI’s claim that he was directly attacking the company, stating that sites that use OpenAI’s API have legal protection, and the company is liable for any damages these sites suffer.
Initially, Xtekky was undecided about taking down the GPT4Free repo. However, after several hours, he revealed that he planned to keep it up and requested OpenAI to file a formal request with GitHub if they wanted to take it down. Xtekky suggested that OpenAI previously tried to pressure him into deleting the repo, but a proper DMCA notice through GitHub would be the appropriate approach.
The possibility of the GPT4Free code being replicated and shared by other community members still exists, even if the original repository is taken down. Regardless of whether GPT4Free existed or not, anyone can exploit unsecured APIs to access GPT4 and GPT3.5. He said the project is being shared and hosted by users everywhere, so removing his repository would have little impact.
On Sunday, Xtekky posted an update on the project’s Discord channel, stating that he is modifying his chatbot, which exists separately from the GitHub repository but serves as a functional demonstration of what is possible with it. We have refrained from sharing a link to the chatbot due to its adult domain name, but the GPT4Free GitHub page includes a link to it.
“We’re moving away from our previous methods and implementing a new system that will guarantee seamless and lawful handling of requests. We’re also excited to announce that we’re rebranding and making general improvements to enhance your experience,” he posted.
Xtekky lastly said that he intends to relocate the chatbot to a different domain and rename it g4f while also modifying the logo to avoid resembling the OpenAI logo too closely. It remains unknown if the repository name will change.