6 Mistakes Leaders Make When Using AI for Storytelling

Have you ever been disappointed by the response AI gave you?

Maybe the writing sounded polished but generic. Perhaps it missed the context entirely. Or maybe it was technically good, but sounded so robotic.

There’s a reason this happens. When we use AI for storytelling, it’s tempting to hand over the hardest part: Here is my topic. Now make it compelling. 

But storytelling doesn’t quite work that way. The experiences that make a story meaningful and compelling belong individually to us. AI can help us shape those things, but it can’t decide them for us. Writing is thinking, and storytelling is meaning-making. They go hand in hand and make us who we are. So do the perspective, judgment, and meaning we bring to those experiences. And if AI simply writes your story but leaves out the meaning-making piece, you’ll get the same thing you see across the internet: overused buzzwords, excessive em dashes or short sentences, the “rule of three” lists, or formulaic phrases like “not just X, but Y.”

As someone committed to the craft of storytelling, with years of experience teaching leaders storytelling as a skill, I began to wonder: What if, instead of writing for you, AI helped you think? What if it helped you build your storytelling skills rather than creating dependency? 

That’s largely the reason we created ella AI. Trained exclusively on over 16 years of my storytelling content and experience, ella AI helps you uncover your own stories and revise them using my storytelling frameworks. It doesn’t replace your thinking- it’s like having me by your side. I wouldn’t write your stories for you. Instead, through questions and feedback, ella collaborates with you to develop a message that is more engaging while helping you preserve and nurture your voice.

Getting useful results from AI requires a different approach than simply asking it to write, especially when it comes to storytelling. Through my years of helping leaders strengthen their voices and their stories (not lose them), I’ve found there are six common mistakes people make when turning to AI.

Mistake 1: Asking AI to write your story

Your experiences are what make your communication memorable and authentic. AI can’t create or generate those experiences for you.

Imagine you’re preparing a leadership presentation about resilience. You could ask:

“Can you write me a leadership story about resilience?”

You’ll probably get a story. But it won’t be your story.

A more useful question to ask AI might be:

“Can you help me identify a story from my career that illustrates overcoming a challenge?”

That prompt changes the task from generation to story discovery. When you begin with something that actually happened to you, AI can then be utilized to help you explore why the experience matters, identify the tension within it, and determine what might resonate with your audience. And usually, ideas start flowing after you’ve found the first one. 

Instead of AI manufacturing a story that it thinks is close enough and compelling, use it to help you discover what you bring to the table: the unique story you already have. 

Mistake 2: Starting with the deliverable instead of the goal

AI requests often describe the format instead of the purpose. These requests sound like:

• “Write me a LinkedIn post.”

• “Write me an email.”

• “Help me create a presentation.”

These are reasonable requests, but they won’t get you far in terms of achieving your goal. Before asking AI to help you communicate something, consider what you want the audience to understand.

For example:

“I want my audience to understand why this project changed how I lead. Can you help me find the right story?”

Now there’s something to work with. The format can come later. First, establish what you’re trying to communicate and why it matters.

This is true whether you’re working with AI or another human being. The clearer you are about your purpose, the more useful the collaboration becomes.

Mistake 3: Waiting until your draft is perfect

It’s easy to think that feedback belongs at the end of the creative process, when the presentation is finished, the email is polished, or a LinkedIn post has been cleaned up. It’s usually then that feedback is sought out. 

But by that point, you’ve already invested so much effort in your approach that changing course can feel difficult. Instead, try bringing AI into the process earlier for feedback.

Bring the messy first draft, your rough outline, the half-written email, the presentation you’re not quite sure about, or simply describe the experience you’re trying to communicate.

Rob Miller, CEO and cofounder of RegAssist, who helped us develop ella AI experienced this firsthand. For years, he had been carrying a story from his time in the Army. He knew there was an important leadership lesson in it, but he had never quite figured out how to articulate it in a way that resonated. When he would try to tell the story, it sounded flat. 

Rather than asking ella AI to rewrite the story, he gave it the story and asked for help.

ella AI asked questions, pointed out where the structure could be stronger, and helped him build tension and clarify the lesson without changing the experience itself.

[Insert Image]

He eventually turned that same story into a LinkedIn post that generated significant engagement. But the more important result was that he had finally found a way to communicate the lesson he had carried with him for years.

As he put it: “I learned a lot from [this experience], but I just never figured out a way to articulate it.”

That’s one reason it can be helpful to bring AI into the process before everything feels finished. Sometimes the rough version contains the very material that makes the final story worth telling, and we just need help structuring it in a way that will help it land with our audience.

Mistake 4: Not seeking out follow-up questions

When it comes to creating an intriguing story, sometimes the best way forward is another question.

Take this example of a recent ella user: Days past his deadline, exhausted after a long day, he needed a short video message for a friend’s milestone birthday. His first thought was to try ChatGPT first. The output was a generic script that lacked any personal touch, which was disappointing for a friendship spanning over 30 years.

He tried ella AI next. Rather than churning out another script, it asked him insightful questions that prompted him to recall a specific memory with his friend. The result was a script that felt meaningful and personal, exactly what he’d been searching for. 

That’s one of the most significant differences between ella AI and other AI models: ella asks you crazy good questions. Being exclusively trained on 15+ years of storytelling experience, these questions are strategically chosen and artfully used to help you discover your experiences or insights that will strengthen your story. 

They can be questions about what surprised you, what you were thinking at the time, what changed, or what you would do differently now.

If you rush past those questions because you’re eager to get to the finished draft, or use general-purpose AI that doesn’t ask them, you may miss the most valuable part of the process. 

Mistake 5: Only utilizing AI when the stakes are high

Imagine waiting until your next keynote to practice good public speaking.

Those high-stakes opportunities, like keynotes, board presentations, or job interviews, are probably not the moments when you want to be figuring out how to communicate more effectively for the first time.

Storytelling, like the skill of public speaking, gets stronger through practice. And practice doesn’t have to mean sitting down to write a formal story; I often encourage clients to try it in the everyday communications that are already happening around you, like a team meeting, a networking introduction, a client email, a LinkedIn post, or a difficult conversation.

These are opportunities to experiment with how to open your story, what details you include, where you create curiosity, and how you connect an experience to a larger idea. The stakes are lower, which gives you room to learn. Then, when the stakes rise, such as an upcoming speech or presentation, effective storytelling feels less like a performance and more like something you know how to do intuitively.

Mistake 6: Using AI that isn’t trained on storytelling 

When it comes to using AI for storytelling, most people turn to a large language model to quickly produce a finished story. However, ella has proven to be an effective, laser-focused tool throughout the entire writing process.

I recently spoke with an experienced AI user who had already built multiple custom AI assistants trained on her own writing. As a Chief Administrative Officer and CFO, she uses AI regularly for executive communication, interview preparation, and performance reviews. In other words, she wasn’t new to using AI as a thinking and writing partner.

So when she tried ella AI, I was particularly interested in what she noticed.

After uploading her resume and personal statement, ella identified stories buried in her experiences and invited her to choose one to explore. Then it began asking questions: Where were you physically? What were you expecting? What made this moment significant? Why did this experience matter?

The process took time. She had to stop and think. But that was part of the point. As she worked through the questions, sharing that ella AI “did draw [my experiences] out” and helped her shape her entire story. 

That is what we hoped ella would do. General-purpose AI is already very good at generating and refining words. We wanted to build something different: a storytelling AI tool that helps you discover the experiences and insights already inside your own materials.

Most people don’t lack meaningful experiences. They’ve lived them, filed them away, and never realized how much those experiences have to teach others.

That’s why ella AI asks questions before it offers answers. What experience is worth sharing? What’s the real tension? What detail am I overlooking? What does this story mean? Why should this audience care?

Those are questions worth asking whether you’re working with a colleague, a coach, or an AI assistant.

The goal isn’t to make storytelling quick and effortless. Good storytelling requires curiosity, intention, and reflection. The goal is to make that reflection more productive.

That’s how ella AI was created. Far from replacing your experiences or your voice, it’s designed to help you explore, nurture, and structure them. It mirrors Leadership Story Lab’s storytelling approach closely so you can communicate effectively, craft messages that stick, and hone your storytelling skills.

The goal is to move beyond producing a finished story and to become a better storyteller. Ready to put it into practice? Try ella AI.