Is AI writing guaranteed to boost a business’ bottom line? For the people who believe the answer is a resounding “yes,” the case is simple. AI tools don’t have to type. They can crank out content exponentially faster than human beings. Some of the most popular AI tools are free.

In terms of driving engagement from readers, the average AI-generated piece might be catching up to the average human piece. In a Villanova-based study published by Cambridge University Press, a majority of participants could not tell whether a short fictional story was written by AI or a human author.

At this point, you might think I’m saying the debate is 100% closed, that using AI tools for writing is guaranteed to boost profits more than human labor.

I actually believe the debate is still open. The potential business impact of a human writer extends far beyond the text they are paid for. With AI tools, the risks and drawbacks extend far beyond quality issues.

The Value Writers Provide Beyond Writing

Writers are also potential business partners and brand evangelists. At The Rauch Review, many of our freelance contributors regularly perform the following services:

• Assisting us with link building
• Promoting their pieces on social media
• Telling their network about The Rauch Review
• Signing up for our email list and getting others to do the same
• Reading and sharing pieces written by other contributors
• Connecting us with business partners and other freelancers
• Donating to my Patreon and getting others to do the same
• Assisting with editing
• Providing their own creative direction and assets for photography and illustrations
• Virtual and in-person event marketing

Sometimes we pay them a little extra for these services. Other times they provide the services of their own volition or insist on pro bono.

Right now AI agents are not advanced enough to provide all of these services. If AI agents do reach this point, we’ll have bigger problems than competition for writing work.

The Value of Real Contributor Pages

On a digital magazine like The Rauch Review, each contributor has a “contributor page” or “author page” that gives more information about the writer and lists all of their pieces. Like most pages on our website, these contributor pages drive ad revenue, affiliate commissions and email signups. From a contributor page, readers often navigate to other pages, further increasing revenue. Contributor pages tend to help with SEO, link building and PR.

Not having human writers usually means losing this page type. There is a way to go 100% AI writing and still have lots of contributor pages, but this strategy is risky and labor-intensive.

To Lie or Tell the Truth: A Lose-Lose Crossroads

When business owners are honest about using LLMs for their writing, they will lose readers and customers who are strongly opposed to AI-generated content or the AI industry.

Another option is the lie of omission: to eschew the conversation by excluding disclosure, bylines, headshots, contributor names and contributor pages. In the case of copywriting, people won’t find a copywriter cited for the project.

This approach doesn’t draw as much wrath from AI haters. When there isn’t any evidence of a human writer, however, a higher percentage of readers will assume AI usage. Accurate or not, AI writing assumptions tend to decrease engagement and increase negative feedback.

The alternative to lies of omission are lies of deception. This process involves brainstorming and creating believable but fake bylines, headshots, identities, contributor pages and more. If you get to the point of making a fake website and social media presence for the fake contributor, you’re likely spending more time than the normal onboarding process for a human.

Once a reader accuses the business of creating fake profiles to mimic human contributors, leadership can be silent, come clean or double down on the lies. All of these responses lead to further distrust.

With human writing, business owners don’t have to take on any of these risks. If someone falsely accuses a human-driven brand or publication of AI content and fake profiles, it’s easy to provide receipts and testimony.

When the AI Bubble Pops, It Will Be Expensive

Free AI tools will not be free forever. There is a limit to how much venture funding they can receive. If they deploy ads and other relatively intrusive monetization methods, a large portion of users will flee.

Many AI companies already charge a fee. Some of these fees, such as token costs, have already been more expensive than human software development labor. The same dynamic could happen with writing.

For the pricing models that are cheaper at the moment, their trajectory is not promising. A team of content writers at Ernest and Young estimated that the cost of a single AI “orchestration” — an action as the result of coordination between models, systems and integrations — is about $1.26 in 2026. They compared this figure to a $0.04 “cost per chat” in 2023. Based on their framing, the cost of an AI task has multiplied 30 times in three years. If this rate of acceleration continues, an AI orchestration will cost at least $37.80 in 2029.

As business leaders increase dependence on AI and use it for increasingly complicated tasks, costs rise exponentially. In the near future, executives might have to choose between abandoning AI and relearning human labor workflows vs. dramatically decreasing the number of orchestrations performed regularly.

AI-Generated Content Tends to be an SEO Risk

SEO has been a big part of my work since 2014. I have dozens of colleagues who are SEO experts. Every time I go on LinkedIn, I see them analyzing data and debating whether website owners should try AI-generated content.

Many SEO experts argue that AI-generated website content almost always has a net negative impact on search engine performance. I’ve seen many posts where freelance SEO people screenshot a report showing the beginning of a pivot to AI content and then a sharp decline in organic search traffic, keywords, etc.

I don’t think Google has a way of knowing how text on a website was created. Even if Google did have a stated preference for human content, it wouldn’t be able to accurately scan whether text is AI-generated.

The trend of AI content degrading websites is real, but the possible reasons behind this dynamic are more complicated than the content itself. AI tools create a temptation for people to produce a vast amount of content in a short period of time. When people indulge in this numbers approach, they often neglect strategic and structural considerations.

Michael Heaton, one of those SEO experts I follow on LinkedIn, did a great job explaining this issue in one of his recent posts. He listed common SEO problems AI tools tend to exacerbate:

• “The page answers a question nobody was asking
• Several pages on the same site answer that same question
• Nothing outside the domain confirms anything the page claims
• The page was published once and never touched again
• No author, or an author with no footprint anywhere else”

Speed, strategic recklessness and a lack of editing increase the frequency of these problems. The more pages you publish, the harder it is to update them in a careful manner.

As Heaton wrote, “When a page took hours of someone’s time and a real invoice, somebody asked whether it should exist. When it takes minutes, nobody asks.”

As for the “author” issue (meaning that the piece has a writer attributed to it), AI-generated pieces don’t have real bylines. It’s possible to create fake human contributors, but then there are the risks and downsides discussed earlier in this article. Or you could be honest and have all of your bylines/contributor pages attributed to ChatGPT, Gemini, Claude, Perplexity, Copilot and so on.

Google rewards the presence of bylines and contributor pages that demonstrate a footprint across the web. With this particular issue, it is accurate to say that Google has a preference for human-written content. Sites with AI content can’t get this benefit.

AI Often Displaces Labor Instead of Replacing It

My editor colleagues have given me insight into how AI has impacted their jobs and career plans. The standard situation is that cutting human writing means saddling a human editor with more work. To finalize AI-generated text, an editor needs a certain level of skill and experience. Even with the job market as horrible as it is, people need to pay this type of editor a decent hourly wage or salary.

To use myself as an example, my last full-time job was as a Director of Content Marketing. I was an editor with supplemental skills in people/content management, SEO, link building, social media, etc. My salary was $140,000 a year, meaning an hour of my time was worth about $67, assuming a 40-hour-a-week schedule.

My boss forced me to take on a project where I hired cheap, low-level writers to quickly generate a massive number of articles. Each article required many hours of editing, sometimes a total rewrite. In another article I wrote, I compared this process to editing AI content, which I later briefly worked on for a freelance gig.

Several of my colleagues tell me that the time it takes them to engineer a prompt and edit is often greater than the time required for the traditional writing process. That time has monetary value. Once the token costs spike higher, will the usage be worth it?

Many of these editors are pivoting careers. They are tired of bosses demanding they increase their hours, with no additional pay, all to edit more AI content, kill more jobs and funnel more profit to the top. When the tokens are too expensive and the editors are gone, how much will employers have to spend on rehiring and rebuilding the human workflows?