Leadership is becoming both easier and harder.

Artificial intelligence has revolutionized how we work, especially over the past year, as it’s transitioned from a secret aid to a welcomed enterprise partner. As a partner, it streamlines work processes, leaving more time for big-picture decisions and strategizing. Each decision, in turn, becomes more impactful. And honestly, it can be overwhelming. Leaders need people around them who challenge their thinking and keep their foot on the gas for innovation.

According to Harvard Business Impact’s 2025 Global Leadership Development Study, respondents are looking for more strategy and creativity from leaders. People now deem skills like leading change, fostering innovation, strategic thinking, and decision making more important than last year.

These insights reveal the expectations people have about business needs. How can leaders ensure they meet these expectations and rise to the occasion? They can either ask people or technology. The catch is, they’re both likely to agree with you.

With people, it’s human nature to agree. Team members get in the habit of wanting to impress their boss, avoid confrontation, and be nice. I’ve seen this firsthand in the two years since I became a CEO. While it can be a nice ego boost, I’ve become apprehensive about any type of perennial support.

WHY “YES” IS COUNTERPRODUCTIVE

“Yes” might be one of the most positive words in the world, but in the business world, it can be counterproductive.

Why? Because it’s overused. We hear it too much, especially in leadership. Sometimes it’s hard to tell when a person is being supportive of a genuinely great idea, or if they’re just afraid to ruffle any feathers.

AI has intensified this concept. Large language models (LLMs) are the ultimate “yes man.” I’ve found they reinforce my perspective by default unless explicitly instructed to counter me. They often double down, even giving me some of my most complimentary feedback.

Even when chatting with colleagues online, it’s so easy to merely react with a thumbs-up emoji over Slack, exacerbating this phenomenon.

People and LLMs have both been trained to agree. But progress stems from challenging that status quo. Leaders’ responsibility now entails building teams that question both human and technology-generated work. Our value lies in asking the nuanced questions that an algorithm can’t.

HOW TO BREAK THE LOOP

Break the loop by finding ways to incorporate dissent. For me, this opportunity arises whenever we do biannual planning at Scribd, Inc. It’s a chance to dig into the nitty gritty, strategize, explore different paths, and think big. And it’s where I try to ensure we don’t fall into the trap of silence after someone asks, “Any questions?”

I don’t pretend to know it all, but here are a few guidelines I’ve found beneficial to encourage this kind of open, strategic conversation.

1. Admit your mistakes. When you’re open, it reassures people that imperfection is okay. Make it clear that you’re not perfect, that you don’t know all the answers, and you sometimes make mistakes. This can prompt others not just to vocalize their own mistakes, but to feel comfortable pushing back and engaging in productive debate as a partner.

2. Foster a culture that treats mistakes as learnings. One thing I love at Scribd is that everyone regularly shares their wins as well as their setbacks, whether in a company all-hands, monthly metrics meetings, or just a quick update in Slack. Beyond the transparency, this allows teams to highlight what they learned when something didn’t go as planned. Ultimately, that’s a win. When people are afraid to fail, they become scared to try anything new. A culture of learning counters this.

3. Bring in the devil’s advocate. Encourage “what if” questions to promote deeper conversations. Model this behavior. After you propose something, instead of closing with “What do you think?”—which can yield a one-word answer—ask a conversation starter like, “What are the potential outcomes here, positive and negative?”

4. Give context. Instead of issuing vague asks that result in employees spinning their wheels to deliver something over-the-top or not aligned with your vision, include the why. Share your intent. Where ultimately do you want to end up? This calibrates the end state, and allows the team freedom to execute.

5. Encourage your people. Build a good team around you. Make them experts in their area. Include them in decisions. Stimulate debate. Get a variety of different types of people. Encourage them to instill this behavior in their own teams.

In today’s world, we all need to work a little harder to break out of our comfortable bubble. Be open to learn, debate, and be wrong. And start looking at disagreement as positive.  

Tony Grimminck is CEO of Scribd, Inc.