AI Generated Content: Privacy Perils and Ownership Rights

Seems like everyone, including writers, business owners and marketers, has jumped onto the AI generated content bandwagon. But before you throw in the whole ensemble, you need to learn about some concerns about privacy, content ownership and copyright. Being aware of these potential issues will help you make informed decisions about using AI generated content.

AI certainly isn’t anything new. If you’ve ever initiated a Google search or interacted with a messaging system on a website, you’ve used AI. We even looked at how Facebook was using AI bots back in 2018.

But AI generated content is a whole new idea. Web writer George Lawton identified at least eight concerns about generative AI systems in his April 18, 2023, article published on TechTarget’s editorial website.

Of those eight concerns, we’ll address four in this post: incomplete and unreliable data, intellectual property privacy violations, questions about who owns the output of a system like ChaptGPT or Jasper.ai and copyright issues.

How Does Generative AI Work?

First, let’s address exactly what a generative AI system is and how it works.

On platforms like ChatGPT and Jasper, computer models have studied as much public information as possible from digital sources–every newspaper, magazine, book, TV show, movie, documentary, doctoral dissertation, website and more–to “mine” or “scrape” information that is then added to the dataset of the generative AI system.

When a user inserts a prompt (input) in a generative AI system, the platform scans all the information it has gleaned and writes an answer (output).

One concern right from the start is that AI systems have only consumed so much data. Developers had to create a cut-off date to stop the data mining process. ChatGPT’s training data stop in September 2021. Any website or digital resource created after that date is not included in the information ChatGPT uses to answer prompts.

I asked ChatGPT: “Who won the 2023 Super Bowl?” and this is the answer I got: “I apologize, but as an AI language model, I don’t have access to real-time information, and my knowledge only goes up until September 2021.”

Who Owns the Output from an AI System?

While users typically have the right to use any AI-generated content how they see fit, there is a question about who technically owns the output. Before you start reprinting output on your blog or on your website, you should probably read the Terms of Service of each platform to learn just what you can and can’t do with that output.

ChatGPT’s terms specifically say, regarding ownership, the user keeps all rights to the input provided. However, OpenAI transfers all of its rights, title and interest in the output (which was generated by ChatGPT) to the user. While OpenAI technically is the owner of the output, the user has the right to use that output, as long as the user abides by OpenAI’s terms.

This may be confusing. But if you’ve ever downloaded a royalty free image from Unsplash, Pexels or Pixabay, you are familiar with the practice. The photographer still owns the image you downloaded, but you have been granted the right to use it in your personal or business materials.

For Jasper, the terms are a little different: “We claim no ownership rights over Customer Content created by you. The Customer Content you create remains yours.”

However, Jasper does retain the right to use any input and to add it to its dataset.

Here’s a quote from Jasper’s Terms of Service: “By submitting, posting, displaying, providing, or otherwise making available any Customer Content on or through the Services, you expressly grant, and you represent and warrant that you have all rights necessary to grant, to Jasper a royalty-free, sublicensable, transferable, perpetual, irrevocable, non-exclusive, worldwide license to use, reproduce, modify, publish, list information regarding, edit, translate, distribute, syndicate, publicly perform, publicly display, and make derivative works of all such Customer Content and your name, voice, and/or likeness as contained in your Customer Content, in whole or in part, and in any form, media or technology, whether now known or hereafter developed, for use in connection with the Services and Jasper’s (and its successors’ and affiliates’) business, including without limitation in connection with modifying, improving, and enhancing artificial intelligence models, as well as promoting and redistributing part or all of the Services (and derivative works thereof) in any media formats and through any media channels.”

In other words, the information you include in your inputs can be used by Jasper to add to its dataset and to help answer other users’ prompts.

What About Privacy? 

Does that mean the information you input is no longer private? According to Jasper’s Terms of Service, it would appear so.

This brings up another concern about privacy and a user’s intellectual property. To keep your private information private, you should keep your personal and sensitive information out of the input box in any generative AI system.

However, you can request the platform to keep your input private. That entails following a lengthy set of instructions on how to make your request in writing and for every piece of input you have entered. There’s no guarantee the platform will agree with your request.

Is Using AI Generated Content Plagiarism? 

Merriam-Webster’s definition of plagiarism is: “to pass off the ideas or words of another as one’s own.”

Does that mean your output in your promotional or informational materials has been stolen from a bot? Can you plagiarize a non-human?

Those are questions the court system will probably decide in the future.

Can AI Hold a Copyright?

U.S. Copyright law specifies that the “office will register an original work of authorship, provided that the work was created by a human being.”

So, no, ChatGPT and other AI generated content cannot be copyrighted.

But can other copyrighted works be used as source data?

Content generative AI systems “learned” about the world by scraping, or reviewing, hundreds of thousands of websites and other information that can be found online, including nearly every book of fiction and non-fiction available online.

In May 2023, several authors, including comedian Sarah Silverman, sued ChatGPT and OpenAI for copyright infringement, alleging their copyrighted works were mined and scraped for data that was included in the AI’s dataset without the authors’ permission.

This appears to be yet another issue that the courts will decide.

What Should You Do?

Make sure you read the Terms of Service for any generative AI system you choose to use. Understand your rights and how the AI system uses the information you provide in your inputs.

AI generated content may be a goldmine for some, but all users need to be informed of their rights and responsibilities associated with using such a platform. Bottom line?

• Use your AI generated content as a starting point.
• Be careful what you use as an input or a prompt.
• Understand there are only so many ways a bot can answer a specific question, so some outputs may be similar to others.
• Rewrite any output and make it your own.

Other things to consider: AI does not understand your brand voice. Another company could pick the copy you produce up and post it online, and duplicate copy is a problem for search engines. Be sure your final content meets your brand voice and is original to your site.

Many writers have said they use generative AI systems to combat writer’s block or to get ideas for content calendars. Check out what we said about avoiding writer’s block and how to develop a blog content strategy in previous posts, and add using AI to your list of options!

What do you think about generative AI systems? Let us know in the comments. And for more information on how Roaring Pajamas can help your business optimize your content, fill out our Contact Us form. We can’t wait to show you what our real-life content partners can do for you and your business.