Copyright Implications of ChatGPT Content

ChatGPT from OpenAI is taking the world by storm. Who owns the ChatGPT output? Can you pass it off as your own? Do you need to acknowledge OpenAI?

Copyright Implications of ChatGPT Content

ChatGPT from OpenAI is a generation-defining technology.

A common question you might be wondering is:  What are the copyright implications of using ChatGPT either directly or as the basis for a published work (like an article on this blog)?

When I asked ChatGPT this question directly, it told me I would likely be violating copyright to use its output:

Q: What are the copyright rules regarding using ChatGPT output as the basis of a blog article?  A: It is generally not advisable to use the output of GPT-3 or any other language model as the basis of a blog article or any other publication without permission from the copyright holder.  In general, copyright law protects original creative works, such as writing, art, and music, from being used without permission. This means that if you want to use someone else's work, such as the output of a language model, you will generally need to obtain permission from the copyright holder or ensure that your use falls under one of the exceptions to copyright, such as fair use.  If you want to use the output of a language model as the basis of a blog article or other publication, you should consider contacting the copyright holder to request permission. In the case of GPT-3, the copyright holder is likely OpenAI.
The content of the above screenshot is reproduced below for easier reading on small screens and screen readers.
[MIKE]: What are the copyright rules regarding using ChatGPT output as the basis of a blog article?
[CHAT-GPT]: It is generally not advisable to use the output of GPT-3 or any other language model as the basis of a blog article or any other publication without permission from the copyright holder.  

In general, copyright law protects original creative works, such as writing, art, and music, from being used without permission. This means that if you want to use someone else's work, such as the output of a language model, you will generally need to obtain permission from the copyright holder or ensure that your use falls under one of the exceptions to copyright, such as fair use.  

If you want to use the output of a language model as the basis of a blog article or other publication, you should consider contacting the copyright holder to request permission. In the case of GPT-3, the copyright holder is likely OpenAI.

The key line, though, is in the second paragraph of its response: "If you want to use the output of a language model as the basis of a blog article or other publication, you should consider contacting the copyright holder to request permission."

So that's what I did.

The OpenAI Terms of Use

As it turns out, OpenAI does hold the copyright to its ChatGPT output.  But–and this is a big "but"–it explicitly assigns to its users (i.e., you) "all its right, title and interest in and to [such] Output."  

From the OpenAI Terms of Use (last modified on December 13, 2022; retrieved on January 3, 2023; emphasis mine):

3. Content
(a) Your Content. You may provide input to the Services (“Input”), and receive output generated and returned by the Services based on the Input (“Output”). Input and Output are collectively “Content.” As between the parties and to the extent permitted by applicable law, you own all Input, and subject to your compliance with these Terms, OpenAI hereby assigns to you all its right, title and interest in and to Output. OpenAI may use Content as necessary to provide and maintain the Services, comply with applicable law, and enforce our policies. You are responsible for Content, including for ensuring that it does not violate any applicable law or these Terms.

In other words, you are free to use ChatGPT's output in whatever way you choose, but you assume all responsibility for it.

Acknowledgement Requirements

You MAY acknowledge the use of ChatGPT content, but it is NOT REQUIRED.

At least, that's the way I interpret OpenAI's current Terms of Use.  Unless I'm missing something (see Disclaimer section below), "all its right, title, and interest" is language that makes providing attribution to OpenAI for its ChatGPT content optional.

Disclaimer

I am not a lawyer.  This is not legal advice.  OpenAI's Terms of Use are subject to change at any time (and may have already changed depending on when you are reading this).  Do your own research.  Contact your own attorney.  

And don't come after me if OpenAI comes after you. 😁

All original code samples by Mike Wolfe are licensed under CC BY 4.0