ChatGPT vs IP

 

DABUS wasn’t considered as the Inventor as it wasn’t the Natural Person.

What about the content created by ChatGPT? The human input or ChatGPT itself or OPEN AI, Who exactly owns that content?

In one of my earlier blogs, I discussed the case of AI being an Inventor citing example of DABUS. In the case of ChatGPT, I decided to converse with it (conversations assembled in several pages). One of the questions I asked it was to write a poetry on a particular subject. It generated an output (a poem), as anticipated. I repeated the question, but now on a slightly different subject. It generated another new poem but this time with few repeated lines taken from the previous ones.

I then ran a plagiarism check on a reliable plagiarism tool, used by Academia as well. The report came neat as no source was cited.

Now, here few questions ensue. DABUS case was different from ChatGPT as in latter, millions of users are concurrently interacting with the remote server where ChatGPT is running. And in case of literature work, it’s difficult, unless any reliable reputable plagiarism tool cites the copied content from the Web, or, there’s already existing prior copyright work (registered or not) that is or be made to be known or is already known or is made available; to discern human vs AI work or plagiarism in general. 

Copyright work comes into legal existence the moment it is created. So, let’s say, an academic article is drafted by ChatGPT, & the human user claims to be copyright holder of that article or thesis (which definitely he/she isn’t, & cannot be), further doesn’t cite or give or name any existing references of any author or of ChatGPT itself; and rather, copyrights that work as Unpublished work in his/her own name.

And as ChatGPT repeats the response, let’s say, the similar article or thesis generated for some other user, that user too copyrights it as an Unpublished work. Now what? The work isn’t available on the Web, thus plagiarism tool cannot process it. The data fed or generated by OPEN AI, again, is not available in the open, & further not shared with any notable plagiarism tools, & furthermore, as not all copyrighted data is available in the data itself nor being taken from the Web nor being compared from the data registered at different Copyright Offices globally; the user who merely entered a question or query, & in response, was being generated an entire article or thesis via ChatGPT; now contends to hold Copyright in the work, that doesn’t even belong to him/her!

So, what’s the solution? Initial solution might be, OPEN AI must open its database of all the generated content & at least share it with the notable plagiarism tools. It must also flag an error if ChatGPT itself generates a selfsame response by running a plagiarism tool on its own server of generated content via ChatGPT. If in case ChatGPT issues a content through Web pages, it must cite that source in the article itself. And furthermore, if ChatGPT is perceived to be as slightly evolving, it must be given the authorship of that content; and Not even to the team behind it i.e. OPEN AI or later even Microsoft.

EQ > IQ. The bridge that connects humans/humanity & AI/Android is EQ & Emotions. The day AI / Androids start exuding feelings (good or bad), that day, they too would no longer be different from a Natural Person.

© Pranav Chaturvedi