I've noticed something odd about AI-generated leadership messages. When they're done well, you can't tell a machine wrote them. When they're done poorly, the robotic corporate-speak or lack of specificity is all you notice.
The message itself disappears. They sound like any generic manager could have written them to any generic team.
Stanford research revealed something crucial. We trained AI to be boring.
Every time a manager accepted a bland response without pushing back, without asking for changes, without interrogating the output, we reinforced the cycle. The AI learned that safe and generic wins.
Most AI-written messages tick the boxes—correct information, proper grammar, appropriate length. But they lack soul.
Researchers discovered that AI models default to safe outputs because of their training patterns, not technical limitations. The more generic content they process, the more generic content they produce. It's a self-reinforcing loop.
The business cost is staggering. Research shows manager impact accounts for 70% of team engagement. When your messages sound robotic, your team disengages. And when communication gaps drive people to leave, organisations face an average cost of AU$50,000 per replacement.
Good AI-assisted communication is specific, engaging, and shows genuine care. It doesn't sound like a template.
Bad AI communication is immediately obvious. It sounds like something pulled straight from a corporate template.
The difference comes down to how you prompt the system. Brief, vague instructions produce brief, vague outputs. Overly complicated prompts overwhelm the AI with too much information at once. But when you give AI the right balance—specific context about your leadership style, your team's composition, the result you want, and the emotion you need to convey—something shifts.
The output becomes invisible in the best way possible.
I developed this approach after watching too many managers struggle with AI tools. Four elements transform generic outputs into authentic messages.
Context means sharing your leadership style and organisational situation. Not just "write an email about a deadline change" but "I'm a collaborative leader who values transparency, and my team has been under pressure lately due to two team members out with sickness. We need to move the Q4 deadline forward by two weeks."
Audience requires describing your team's composition and preferences. Not "send this to the team" but "My team includes three early-career staff who need clear direction, two senior members who prefer autonomy, and one person working remotely across time zones." Individual people with different needs, not a homogeneous group.
Result specifies the desired outcome and behaviours you want to encourage. Not "get them to accept the change" but "I want them to understand why this matters strategically and why their input is critical to success. They need to feel confident they can deliver something that wouldn't be possible without them, and know that meeting this objective means they'll meet their bonus and pay rise criteria."
Emotion identifies feelings to evoke or avoid. This is where most managers fail. They ignore emotional nuance entirely. Not just stating the facts, but "They'll likely feel frustrated about another deadline change. Acknowledge the pressure they're under, convey confidence in their abilities, and avoid creating panic about the timeline."
A vague prompt produces corporate text. An overly complicated prompt creates confusion. A CARE-enhanced prompt generates genuinely human messaging. Unlike approaches that ask you to generate five options and pick the best, CARE gets it right the first time. Time-poor managers need efficiency, not more decisions to make.
I walk through each element in detail in my practical guide, including templates and step-by-step application for different scenarios. Get the guide here.
The truth is, most managers don't do this well. They lack both communications expertise and AI training—promoted because they're brilliant at their job but never taught how to communicate across a team in a way that inspires everyone to feel part of something special.
But here's what matters. AI should amplify your communication, not replace it. You provide the authentic care and understanding. AI provides structure and efficiency.
Harvard Business School research by Professor Prithwiraj Choudhury reveals that employees rate messages as less helpful when they believe AI wrote them, even when the content is identical. The perception of care matters as much as the care itself.
That's why specificity wins. When your prompts include context, audience, result, and emotion, the AI output reflects your genuine concern for your team. It becomes invisible because it sounds like you.
We trained AI to be boring by accepting generic outputs. But here's the good news: we can train it to be better by being more specific about what matters.
When you give AI the context, audience, result, and emotion behind your message, you're not just getting better outputs. You're finally getting the communication support that senior executives have always had — adapted for your team, your situation, your leadership style.
That's what every manager deserves. And now it's within reach.
Get the guide here.