Category: Articles

Just a few months ago, the European Journal of Innovation Management published an article about AI adoption in digital content marketing. Talking about the ethical and legal implications, the researchers raise a question: who should hold the authorship of content generated by AI and then perhaps revised by a human? They say, “This ambiguity can lead to complexities regarding ownership of the generated content and potential infringement of third-party copyrights.”

It’s interesting that some academic papers look at the issue from one side only. In this case, if a marketer generates an article and takes ownership of it, it won’t be fair to AI companies. As a counterargument, I would definitely include the statement from The New York Times, which says, “All of our copyright claims will continue against Microsoft and OpenAI for their widespread theft of millions of The Times’s works, and we look forward to continuing to pursue them.”

Then, to make my argument stronger, I would further include another statement from eight different publications, made in 2024: “We’ve spent billions of dollars gathering information and reporting news at our publications, and we can’t allow OpenAI and Microsoft to expand the Big Tech playbook of stealing our work to build their own businesses at our expense.” OpenAI’s response to all these statements is based on one foundation: fair use.

Personally, I don’t take sides in these disputes because the problem is truly complex, and while I understand the concerns of journalists, I also find OpenAI’s explanation very reasonable. The reason I brought these up was simply to question the academic discussion about the authorship of AI-generated content. If we go deep into this, we will find ourselves in a rabbit hole with no just answer.

The research mentioned above cites Formosa et al. (2025) as offering one solution to the issue: “hybrid authorship” and “shared responsibility.” So, assuming marketers should share authorship of AI-generated content with AI, I invite you to look at this matter from another dimension.

The man who has been in the headlines of global news these days, British politician Andy Burnham, has spoken openly about the need to regulate social media, and another politician, Wes Streeting, pushed the conversation further by talking about the need to hold AI companies (he mentioned Grok) accountable for the output of their language models. We can see similar conversations elsewhere too.

People are worried that if this really becomes policy, it will make the existence and operation of AI platforms in the UK practically impossible. Why?

ChatGPT receives 2.5 billion prompts every day. This means that by the time you blink, it processes around 29,000 inquiries. Grok, on the other hand, has about 117 million monthly active users, while daily output volumes are not published by xAI.

Having said this, how do you imagine an AI company being accountable for every output its chatbot provides in such huge numbers every day? It is logically impossible. But if we look at the technical side of this, these companies have a deeper reason to reject any such demands. AI assistants do not provide pre-written content that can be verified or approved in advance.

They literally generate outputs by predicting the next word, and then the next word after that. To be fair, a company should be accountable for system design, safety measures, and known risks, but not for every single output its software produces. And that’s what some critics are worried about, considering that demanding the impossible may eventually lead to banning AI assistants.

Circling back to the co-authorship proposals by academics, if we assume that a marketer generated content with AI from ideation to writing, then when they publish it and something goes wrong because of misleading information, factual inaccuracies, etc., who bears the responsibility for removing it, explaining the mistake, apologizing for it, and maybe even writing a resignation letter? The marketer.

And so, here is the logical analysis based on this reality.

For AI platforms to share authorship of content, they also need to share responsibility for it. And since we saw above that it is both logically and technically impossible for AI companies to be responsible for every single output their language models produce, it is fair for marketers to hold full ownership of every piece of content they generate with AI and choose to publish. And this is not a favor to them, but a weight of responsibility.

In the 1st episode of the 7th season of The Office, Michael Scott hired his nephew as an assistant. And as you would expect from Michael Scott, it was a bad hire. This episode shows how naive he was and how bad he was at managing people. But this episode also shows how great a marketer he was.

Everyone in the office complains about Luke because he is bad at carrying out his duties. Very normal: remember, he was Michael’s nephew, and that was the only fact Michael could keep secret for quite a while. Once he reveals everything, the employees wonder: Why wouldn’t you tell us? And this is where Michael proves he is a marketing genius.

“Because I wanted you to come to me,” he explains. “And say, ‘Wow, he is so great.’ And I was gonna say, ‘Well, it’s in the genes.’ And I was actually gonna be wearing jeans.”

Interestingly, the official YouTube clip cuts out that part of Michael’s speech, which would have come after 1:44.

You know who else has great jeans? Sydney Sweeney.

In 2025, American Eagle Outfitters (AEO) launched a campaign featuring the 27-year-old actress. She speaks about her hair, eye color, and personality, saying that “genes are passed down from parents to offspring.” Then she uses a pun, stating that her jeans are blue. The ad concludes: “Sydney Sweeney has great jeans.”

This is exactly what Michael Scott was thinking about but never had a chance to do. And maybe that was for the best, because after American Eagle did it, a huge backlash started against the company and even Sydney Sweeney.

This is a very interesting case study in marketing for two reasons. First, it shows that audiences have changed dramatically in the past 15 years, and what they would have laughed at back then may now seem insulting. Second, it may still work.

About the first point: audiences change. That’s undeniable, but the marketing application is slightly deeper: no messaging is valuable or creative in isolation, and everything must be examined in the current context, even if it is borrowed. I found no records showing that anyone at American Eagle was inspired by Michael Scott, but the similarity is too close to ignore (come on, the series was streamed for 57 billion minutes).

People’s criticism was multifaceted and included ideological accusations such as “white supremacy,” feminist concerns such as the sexualization of women, and so on. Some of these arguments felt overstretched, especially because American Eagle has worked with diverse public figures before, but they still show how quickly a campaign can lose control of its intended meaning.

But this doesn’t matter in the age of social media and user-generated content, because the decoding of brand messaging ultimately belongs to internet users. If they decide the campaign is wrong, then it is, at least for them. And I don’t mean that a brand should be afraid of being misunderstood and become a people-pleaser.

My second point was about the fact that, despite being misunderstood, the campaign worked. It generated 40 billion impressions on the internet. This includes people commenting, liking, sharing, and so on. But one thing I strongly dislike is presenting social media metrics as business achievements without talking about their actual impact on the business.

Reports vary, but the official figures from American Eagle’s chief marketing officer, Craig Brommers, show that after the campaign, the brand acquired 790,000 new customers, saw a 2% increase in revenue through the end of that quarter, and had a stock jump of 25%.

I have other observations too that show that Michael Scott—or whoever wrote the episode—was a great marketer, and maybe I will write about them later. That’s a personal thought. But the underlying fact is that he predicted a 2025 advertising trend 15 years earlier.

Although the audience has changed since then, some things never get old anyway. Just like a good pair of jeans.

I am not an AI advocate, but I consider myself an AI power user. Still, just because I have been mastering this technology, I don’t want to convince the world that everything must be replaced with algorithms. Most things shouldn’t.

The Economist published an article about the use of AI in various industries. It provides data about the number of mobile apps before and after the emergence of vibe coding. I have never been an avid ebook reader, so maybe I am lucky there: according to the data, their number has tripled since 2022, when some writers started putting their prompting skills to use.

Anyway, all this is expected, but one statement in the article worries me: a lot of people actually prefer AI-generated text, music, and software. What this means to me is that, as consumers, we often care about the final result while being indifferent to the soul of the work. People say they have nothing against AI-generated music because, first, they hear no difference, and second, they like it. Pretty logical, right?

If you ask me, the blame is not on OpenAI, Google, or Anthropic. It’s Spotify. And the shuffle culture it has successfully popularized. I’m not saying Spotify started it. Artists themselves have spoken against this, and I remember Adele’s public statement on social media: “We don’t create albums with so much care and thought into our track listing for no reason. Our art tells a story and our stories should be listened to as we intended. Thank you Spotify for listening.”

This was the only request I had in our ever changing industry! We don’t create albums with so much care and thought into our track listing for no reason. Our art tells a story and our stories should be listened to as we intended. Thank you Spotify for listening 🍷♥️

— Adele (@Adele) November 21, 2021

Eventually, we came to the point where many of us have forgotten how to listen to music, or what to listen for in music. I am not even talking about the composition of entire albums, but individual songs. Many may listen to “Regrets, I’ve had a few. But then again, too few to mention,” and never realize that, through his performance, the singer shared someone’s experience, told a story, showed vulnerability, and maybe even revealed some concealed pride in his personality. Music is story, emotion, psychology, and much more.

And so, when a lived life turned into a ballad becomes nothing more than background noise, people stop caring about the creator. I just need something that makes me feel good, and if it’s AI-made, no difference.

The same goes for books.

The AI policing we see today will never solve any of these problems. I know some readers of this article will, just out of curiosity, scan this text through an AI detector to check my honesty. Don’t. Because AI policing will always punish the wrong suspect.

The indifference toward music started when generative AI was still science fiction. And had it not been for that indifference, no one would worry about AI-generated music today, because it wouldn’t have a significant audience.

In any text, product description, press release, or official statement, it’s not the AI “score” that makes the difference. Marketers need to think about the big picture, going beyond structural details. When the message conveys the right idea, provides something valuable, and speaks to the reader, the customer, or the business partner, most people will not care whether AI was involved in creating it. And in most cases, they shouldn’t have to.

We can’t change our perception of art, literature, and marketing by policing AI. Let it be. Every creator simply needs to ensure that their work has a human reason to exist.