It used to be that to stand out at work, we needed to be fast, polished, and agreeable. In the age of artificial intelligence, all of that gets turned on its head. Now, anyone can produce anything quickly with the appearance of poise and tact.

For nearly two decades, I've coached CEOs and senior executives on how to communicate, collaborate, and stand out as leaders. But over the past few years, I started noticing something troubling: the very qualities that once made people successful at work were becoming easier to automate. If everyone has access to the same intelligence, what will actually make someone unforgettable?

For my forthcoming book, " Use Your Brain " I interviewed CEOs, scientists, artists, parents, and students across five continents. As someone who was becoming increasingly uncomfortable with my own AI use, I sought to understand what separates people who use AI to sharpen and improve their thinking from those who surrender to it and lose their point of view.

What I found was that the people who retained their autonomy and kept their work memorable had developed these counterintuitive habits.

1. They slow down at the start I've found the best critical thinkers don't rush. When using any AI tool, they employ what psychologist Sumona De Graaf calls a "Think Sandwich" — Think, AI, Think. To build your own "Think Sandwich," take five minutes to answer three key questions before you prompt. What do I believe about this? What problem am I actually solving? What would a bad outcome look like?

The first one forces you to develop your own point of view before AI gives you one. The second one keeps you from getting a brilliant answer to the wrong question. And the third one makes you consider the risks and consequences that a polished AI response might overlook.

Without them, you risk solving the wrong problem, accepting assumptions you never questioned or mistaking a convincing answer for a good one.

After you answer these questions, then consider using AI to review your research or first draft. You can also invite it to attack your argument or poke holes in what you wrote. But I would also get a gut check from another human being who knows the context that you are working in.

2. They are less polished I had a coaching client, a newly promoted senior vice president at a large company, who had started letting an AI assistant draft his email replies.

A few months later, he received his annual review from his direct reports. He told me that one colleague wrote that it had become hard to tell where he stood on anything. Another said that he sounded "polished," but didn't frame it as a compliment.

It used to be that polish signaled credibility. Now with the proliferation of AI tools, that polish can seem cheap or empty. But being less polished doesn't mean leaving bad grammar and typos in your work. It can look like: Starting an email or text message with a detail only you could know. Putting your stance on a subject in the first sentence, rather than clearing your throat with small talk. Naming one thing you're unsure about.

This week, before you send your next important e-mail, ask yourself: "Could anyone else have written this?" If the answer is yes, look for ways to be more specific. Authenticity will resonate more than perfection.

3. They go looking for disagreement When everyone has access to tools that validate their thinking, we risk creating workplaces where assumptions go unchallenged and bad ideas sound brilliant. That makes healthy disagreement at work more important, not less.

The people who stand out will be the ones who are willing to challenge the thinking in the room, including their own. Here's how to go looking for disagreement in a constructive and helpful way: Ask questions instead of delivering verdicts. Start with what you agree with and follow that to where you're stuck. Assign a designated skeptic in meetings to normalize pushing back.

Critique your own work first. When you share a draft with colleagues, say, "Tell me the weakest part." These habits make disagreement feel less like a personal attack and more like a shared effort to get to a better answer. Questions lower defensiveness, finding common ground builds trust, and inviting criticism makes it safer for others to speak up.

4. They protect their independent judgment When I ask executives who they believe are the most at risk of handing their thinking to AI, they usually point to Gen Z or junior employees. The irony is that I see the opposite in my coaching work. This is because executives' roles have shifted from doing the work to reviewing it. Now AI makes that review process even easier, serving up polished summaries, recommendations and decisions in seconds.

The danger is that leaders who spend all their time just approving answers then struggle to identify biases in their team's work. This leads to A-level work on the surface, with C-level understanding of the issue at hand.

The higher you rise, the more consequential that loss of independent judgment becomes. To intentionally protect their thought processes, I recommend that leaders: Do one meaningful piece of work from scratch every week.

Explain the core logic of anything AI-assisted out loud, without a screen or written notes from anyone else, before presenting it.

Ask themselves the question they'd normally ask their team: "Walk me through how you got there. What assumptions is it based on? What alternatives were rejected? What evidence would change your mind? What are you missing?" If you can't answer those questions without looking at the AI-generated summary, you may be approving a conclusion you don't fully understand. Have team members give 90-second presentations where they explain, defend and recall their point of view.

5. They choose to have the live conversation If your message contains emotion, conflict, disagreement or requires a hard decision, pick up the phone or organize an in-person meeting to have a real-time conversation. Don't just show up to deliver your message. Start by asking a question you genuinely don't know the answer to and listen without planning your response or rushing to a bot.

Pay attention to what's not being said, whether that is hesitation, a change in tone, or a concern someone seems reluctant to raise. Before concluding the meeting, make sure the other person feels heard by echoing what matters most to them.

I've found that the conversations people remember aren't necessarily the ones where you said something brilliant. They're the ones where you made someone feel understood, challenged their thinking with care or had the courage to say what needed to be said.

AI can help you craft the perfect message, but it can't replace the trust that you personally build in the moment.

Erica Dhawan is a leadership expert, keynote speaker, and Wall Street Journal bestselling author of "Digital Body Language" and "Get Big Things Done." Her new book, "Use Your Brain: Think Deeper in a World on Autopilot " is out Jan. 12, 2027.

Take control of your money with CNBC Select CNBC Select is editorially independent and may earn a commission from affiliate partners on links.

7 simple inflation hacks to save money every month Most Americans spend too much on housing. Here are 3 ways to save on your rent or mortgage How to save on hotels and flights ahead of the most expensive holiday travel season in 10 years How long does debt relief take?

14 valuable credit card welcome bonuses ending soon: Earn $1,000+ in travel with the top offers