The assertiveness framework that works on AI too

You already know that assertive communication transforms your career.

When you're direct with respect, colleagues listen. When you replace "sorry" with "thank you," people see you as a leader.

But here's something fascinating I discovered recently...

The exact same communication principles that make you influential with humans also make you dramatically more effective with AI.

Think about it. You've been learning how to prompt ChatGPT, Claude, or other AI tools. When you get inconsistent results or vague responses, what's your first thought?

"This is useless."

"AI is just overhyped."

"I'll stick to doing this the old way."

So you stop using it for important work. Maybe you use it occasionally for basic tasks, but when colleagues start talking about how AI is transforming their workflows, you're quietly thinking it's not living up to the hype.

Meanwhile, you're falling behind without realising it.

But here's what I want you to know: It's not that AI is useless. You just haven't been taught how to communicate with it effectively.

Recent studies show that the tone, clarity, and directness of your prompts dramatically affect the quality of AI responses. Just like when you're talking to your colleagues.

The hedging language that undermines your authority with humans? It does the same thing with AI.

The clear, confident communication that makes you a respected leader? It gets you better AI results too.

Here's what studies found

I know this sounds almost too simple to be true. But researchers around the world have actually studied this.

Multiple recent studies show that the emotional tone, politeness level, and clarity of your prompts dramatically affect AI performance. The parallels to human communication are striking.

Here's what researchers discovered:

Study #1: Your emotional state transfers to AI

A 2025 study conducted by a Washington-based AI startup looked at something called "sentiment propagation" in prompts. In simple terms: the emotion in your language affects how accurate the AI's response is.

When you use negative language (frustrated, uncertain, apologetic), the AI focuses on dealing with that emotion instead of giving you the best factual answer. Result? Shorter, less accurate responses.

When you use overly positive language (excessive flattery), the AI sometimes exaggerates.

But here's the sweet spot: neutral, direct language gets you the most accurate results.

Study #2: Politeness gets you better results

Another study from October 2024 at the Waseda University in Tokyo tested how politeness affects AI responses. The findings?

Impolite prompts led to poor performance. But here's the interesting part: when prompts were more polite, the responses had more length and detail.

The researchers explain it like this: "In human culture, a highly polite environment makes people more relaxed and willing to express their true thoughts. The behaviours of GPT-3.5 and GPT-4 may precisely reflect such human behaviours."

Study #3: Clear language reduces confusion

A third study by researchers across universities in the US, India and Bangladesh found that precise prompts enable AI to generate content that's more aligned with what you actually need. Vague language creates uncertainty. Clear language guides the AI toward the correct response.

The research shows: Prompts that use clearer and more specific language tend to generate fewer hallucinations.

Translation? Stop the "maybe" and "I think" language. Be direct.

Why your prompts aren't working

Here's an example prompt you might use:

"Can you help me understand ML algorithms? I am unsure how they apply to fraud detection."

Seems reasonable, right?

But let me break down what goes wrong:

"Can you help me understand ML algorithms?" Too broad. There are dozens of ML algorithms - which ones matter for your situation?

"I am unsure how they apply to fraud detection." This signals uncertainty instead of directing the AI toward your specific needs.

No context about your role, project, or what you need to do with this information. The AI has to guess whether you're a beginner or expert, building something or just learning.

It's like walking into a meeting and saying, "Can you help me understand our strategy? I'm unsure how it applies to our project."

You wouldn't get great results with humans either.

Your assertive AI prompting framework

Here's the framework that consistently gets better results from every AI tool:

1. Acknowledge the task and request focus

"This is a complex/important task, so I'd like you to stay focused and precise in your thinking."

This sets the tone. You're signalling that this matters, and you expect quality.

2. Share your purpose or context
 

"I'm [describe your role/project]. I'm working on [brief context: e.g. a new programme, client pitch, article, research study]."

Give the AI enough context to understand your world, but don't over-explain.

3. State the goal clearly
 

"My goal is to [e.g. understand competitors / draft a curriculum / map out current thinking]."

No hedging or "maybe." Just clear direction.

4. Suggest a task or output

"I suggest you [e.g. perform research, write a summary, evaluate options]. Here's what I'd like covered: [bullet points or numbered list]."

Be specific about what you want. Structure helps.

5. Ask for clarification or feedback
 

"Do you need any clarification before starting?" "Does this make sense? Anything missing from my brief?"

This opens the door for the AI to ask questions, just like you would with a human colleague.

Watch this transform a real prompt
 

Here's how to transform that weak prompt I showed you earlier using this framework:

Original weak version: "Can you help me understand ML algorithms? I am unsure how they apply to fraud detection."

Assertive version: "This is a strategic research request - please stay focused and concise in your output.

I'm a data scientist at a fintech startup. I'm working on improving our fraud detection system using machine learning.

My goal is to understand which algorithms would be most effective for real-time fraud detection.

I suggest you analyse the following:

1. Which ML algorithms are best suited for fraud detection?

2. What are the trade-offs between accuracy and processing speed?

3. What data preprocessing steps are most critical?

4. How do leading fintech companies approach this problem?

Do you have any questions before you begin? Let me know if more context would help."

The result? The AI gives you a detailed, structured analysis that directly addresses your needs. No more vague responses.

Why this matters for your career

Think about it: we're all using AI more and more in our work. Reports, analysis, research, coding, writing.

The people who master AI communication will have a massive advantage. They'll get better results faster, look more competent and save hours of back-and-forth.

And the skills that make you better at prompting AI? They're the same skills that make you a better leader:

• Providing context

• Setting expectations

• Being respectful but assertive.
  
  

Your next step

Pick one AI task you do regularly. Maybe it's writing emails, analysing data, or researching competitors.

Use the 5-step framework above for your next prompt. See what happens.

And remember: the same assertive communication principles that work with your colleagues work with AI too.

Be clear, direct and respectful.

You'll be amazed at the difference.