The 3 AI Conversations Every Leader Should Be Having - Forbes

With Anthony Gostick

Imagine it’s 1912. You're behind the wheel of a Stutz Bearcat, one of the fastest machines the world has ever seen. You’re testing your new automobile’s limits. At nearly 80 miles per hour, the engine is roaring, the wind is in your face, and you are moving down your town’s dirt road at a speed that would have seemed like science fiction just a few years ago.

There are no traffic lights, no lane markings, certainly no speed limits. You open the throttle and come flying around a bend to find a farmer and his horse-drawn hay wagon directly in your path.

Uh-oh.

Now, the farmer pulling his cart did nothing wrong. And as for you, well, you had every legal right to be going that fast. But that didn’t stop the collision or resulting chaos.

Which brings us to AI.

Organizations suddenly find themselves with a much more powerful machine at their disposal. The potential is exhilarating. Yet where this road leads, or what dangers lie around the next bend, is far from clear. The only thing most people seem to agree on is this: wherever we are headed, we sure are getting there fast.

Many leaders we work with have spent the past years experimenting with and learning about AI, watching what it can do, and trying to figure out what it means for their teams. It’s a mixed bag.

Some days working with AI can feel like arguing with an overly confident intern.

You: "I just told you not to do that, why did you just do it?"

Intern: "Oh, you're right. You did tell me not to do that. Would you like me to ignore your next request too?"

Other days, AI can genuinely feel like magic. In the past year, one colleague of ours has written five books. In the previous decade, he wrote one. A small company we work with has rebuilt their entire backend from scratch, work that would have taken a development team the better part of a year. The bill came to almost nothing, and the work was mostly done by non-technical people.

That’s the future most leaders are chasing: capable people doing more for less. And the excitement of that is understandable. AI has already proven its ability to revolutionize industries. But in order to succeed long term, smart leaders are starting to weigh the risks of adoption carefully before rushing ahead.

When individuals, teams, companies, and entire industries suddenly start moving this fast, infrastructure must be built and rules put in place. Yet many leaders are still so excited about the speed that they are not slowing down long enough to ask, “What’s around the bend?”

The same technology that is capable of accelerating drug discovery, powering scientific breakthroughs, and helping your team get more done in a week than they used to in a month can also create system vulnerabilities, spread sophisticated misinformation, enable surveillance, and automate cyberattacks at an unprecedented scale.

The upside of AI is real, but so are potential downsides. Which is why leaders need to get serious about defining the rules.

The Conversations No One is Having

Right now, most organizations are treating AI as a technology implementation project. They’ve invested in platforms from the big AI providers and are discovering just how much this tool can do for them. But most leaders are still skipping the harder conversations, often out of fear of being left behind: What are the risks here? Where are the limits? And perhaps most importantly: How does any of these tools really even work?

As for employees, they are watching and waiting. They know AI is coming, but they are asking where they fit into all of this. They know that some jobs will evolve, some will become more productive, and some will disappear. That happens with every new technology. They can handle that hard truth, what they can’t handle is being left out of the conversation entirely.

What the world needs now are leaders willing to hold honest conversations with their teams and communities to find the best road ahead.

None of this requires developing a ten-year AI strategy or bringing in a team of expert consultants. What we need are leaders willing to start talking.

Where to begin:

1. The Adoption Conversation.

If leaders engage in conversations around AI’s use, the first question most leaders ask is: what can AI do for us?

Smart leaders are taking a different approach. They know AI is going to help. Rather than rushing to implement, they are working with their teams to develop a clear stance on what their organization will and won’t use it for. They’re look at costs to build specific tools internally and weighing that against provider-based options (or if they want to wait for better tools to emerge).

They aren’t just using AI, they’re collaboratively creating a crystal clear roadmap of when and why to adopt these new tools. That includes the most overlooked part of the AI adoption process: conversations about how not to use AI. Smart leaders are drawing lines now, deliberately.

An AI adoption strategy like this is not intended to exist as a policy document buried in an HR portal that no one reads. It should be a living, breathing way of guiding the organization’s AI implementation strategy.

Unfortunately, most leaders are skipping these conversations and jumping straight to implementation out of fear they’ll miss out. Don’t let fear keep you and your team from having a solid AI adoption strategy.

2. The Informed-Use Conversation

Once leaders have established where AI fits in the organization and how it will be adopted, the next conversation is about you will help people learn to use it and take ownership of it.

Every employee can today have access to an intelligent assistant for the price of a few cups of coffee each month (or almost free if hosted locally). The organizations that pull ahead won’t just buy AI tools and hand them to their team members, they’ll customize tools, build cultures where people are encouraged to experiment safely with AI, and work to continually adapt to find new ways this technology can improve or assist their work.

That doesn’t mean every company should let their people start writing code, and it doesn’t mean people should drop the responsibilities of their current jobs and focus on their newfound “AI Art” skills. It does means that people need to learn the basics of how these new tools work. Just as we expect professionals to know how to create spreadsheets, build slide decks, or run search engines, we should expect them to understand how to properly prompt, challenge, and work with AI.

For leaders, that raises several questions: How are we going to help employees learn? Who is identifying genuine cases of better ways of working and sharing them across the organization? Are we creating safe opportunities to experiment and what are the guardrails? Which AI tools are safe, and which introduce unnecessary security or privacy risks?

Some organizations we’ve studied are benefiting from creating an AI playbook, an "AI Bible,” that explains which tools are approved; how data should and shouldn't be handled; and examples of effective prompts, preferred workflows, and the organization's principles for responsible AI use. Without shared guidance like this, employees may each try to invent their own approaches, creating unnecessary inconsistency and risk.

Informed use like this starts when leaders themselves become informed. As a leader, you don't need to get a degree in computer science, but you do need a working understanding of what today's AI systems can and cannot do for your teams. That means understanding concepts such as prompting, tool calling, memory storage, data privacy, and when self-hosted or enterprise AI solutions may be appropriate. Every leader is going to need to become an informed guide for their team.

3. The Value Versus Utility Conversation.

Good governance of AI comes down to this big idea being discussed before every implementation: Will using AI here actually make things better for a key stakeholder? That stakeholder may be your customers or it may be your employees. Ideally, it's both.

Yes, every AI initiative should consider the bottom line (savings) or top line (revenue). But just because AI can do a job, it doesn’t mean it should. Especially if your employees or customers become guinea pigs.

Consider customer service chatbots. Here, AI may reduce costs and seem efficient, but it may also create frustration and erode trust by trapping valuable clients in endless loops before allowing them to speak with a human. If your business can’t afford to lose customers, it may not be the right call to try this out on them without plenty of beta testing and a careful, staged roll out.

Now, contrast that with AI that equips customer service representatives with a summary of a customer's history and likely products and solutions they’ll need before the phone is even picked up. Then, the technology is helping both employees and customers rather than getting in their way, and rapid implementation probably makes a lot of sense.

AI that eliminates repetitive or tedious work, emerges insights employees couldn't have found on their own, or gives customers more of what they need will create value that everyone can benefit from.

Great leaders don't ask, “Can we use AI here?" They ask, "Will our people or our customers be better off because we did?”

Want to Thrive in the Future?

The cultures that thrive going forward will have leaders who model the right kinds of behaviors. So, be honest about what you’re still figuring out, what you’re working on, and what you’re interested in. Acknowledge that nobody in your organization has been here before, and get people involved in conversations about their futures.

The potential of AI is real and exciting. We can and will move faster than we ever have. The leaders who navigate it won’t just press the gas pedal to the floor and hope for the best. They’ll decide in advance how to grow alongside this new technology while still being mindful of the road ahead.

No one has all the answers. But leaders who survive this transition will be the ones willing to slow down and start writing the rules today.