I like to think of Artificial Intelligence as simply math applied to data. When the data is your own, such as your original writing, AI serves as a tool to modify what you’ve already done. However, using AI to modify other people's work means ethical and legal issues around copyright and ownership.
This is the third post in a series on AI and marketing I am writing myself, (with AI help to generate headlines) this month. The other two posts are Generative Artificial Intelligence and the Future of Marketing and Unlocking the Potential of AI for Use with Your Own Content.
Using AI tools like ChatGPT to write content from data sourced on the web and then post that content as your own presents a variety of issues. Let’s look at a few:
AI can be wrong
Whenever you rely on technology to create content, you run the risk of having inaccurate data.
For example, I found this article in TechBullion: Strategic Marketing Requires Lifelong Learning: 3 Marketing Books To Help Your Brand This Year.
I ran the article through GPTZero, a free tool that analyzes content and provides details for how much of the document is written by AI. The results indicated that yes, it was likely written by AI.
I included this example for another reason. The second book listed in the article is Marketing in the Age of Pervasive Media by David Meerman Scott & Brian Halligan. Hmm... Interesting… My friend Brian (HubSpot co-founder and executive chairman) and I didn’t write a book with that title! Clearly the AI messed up and the “author” of the TechBullion article never checked it for accuracy.
AI uses other people’s work
When AI is using data on the public Web, it is using other people’s work to generate content. This can be problematic if discovered. Let’s look at an example.
I’m the pioneer of Newsjacking. I wrote a book with that title, and I have the URL newsjacking.com which is typically the top search result. I’ve written many articles and given many speeches on the topic, and I’ve been a guest on countless podcasts and video interviews talking about newsjacking.
When I asked ChatGPT from OpenAI “What is Newsjacking?” this was the answer:
“Newsjacking is a marketing strategy in which a company or brand tries to align itself with a current news event or trending topic in order to capture public attention and promote its products or services. The goal is to insert the brand into the conversation surrounding the news event, thereby increasing its visibility and relevance. This can be done through various means, such as social media posts, advertisements, blog articles, and more. The key to successful newsjacking is to be quick, relevant, and respectful.”
No, this isn’t direct plagiarism. But I can smell my work in this paragraph.
The main image I used for this article was generated by OpenAI’s DALL-E 2 using the prompt “graffiti image of a person reading a book”. Yes, it was created by using data from the public web.
AI and copyright law
My buddy Mitch Jackson is a lawyer who has been studying the legal aspects of AI. He wrote an excellent LinkedIn article: A Quick Overview: Navigating the Gray Area of AI and Copyright Law. Mitch allowed me to summarize the three important points in his article:
Human authorship - Refers to the attribution of a creative work to a natural person as the creator or originator of the work. When a human creates a work, it's clear who the author is. But when an AI system creates something, who should be considered the author?
Originality - Copyright law requires that a work be original to be eligible for protection. This ensures that only truly unique and creative works are eligible for protection. But what does that mean for a work created by an AI system?
Fair use - Copyright law allows for the use of copyrighted material in certain circumstances and without permission for the purpose of criticism, commentary, news reporting, teaching, scholarship, or research. But what happens when an AI system uses copyrighted material to create a new work? Is that a fair use, or is it copyright infringement?
I suspect that there will be cases that will soon go all the way to the U.S. Supreme Court to figure out these important questions.
Moral and ethical considerations of AI
My friend Paul Roetzer, founder of the Marketing AI Institute, has developed The Responsible AI Manifesto for Marketing and Business.
The Manifesto opens with, in part: “AI should make us better people, professionals, and organizations. However, this will not happen without a continuous focus on the responsible application of AI across all business functions. We have to be willing to have the hard conversations now so that we do not ruin what can be one of the most transformative technological shifts in human history.”
The manifesto can be used under a Creative Commons license as a starting point for companies interested in developing AI guidelines.
Disclosure: I am investor in and an advisor to the Marketing AI Institute.