There’s no shortage of hype around the potential for AI to transform the workplace. A recent McKinsey report compared the tech to the birth of the internet and the arrival of the steam engine.

But, its reality is still taking shape. AI adoption is inconsistent at most organizations, workers have varying levels of interest, and there’s often a difference between AI buzz and its practical application.

CarGurus, an online marketplace for buying and selling cars, is one company trying to bridge that divide. Last October, it launched AI Forward, a 20-person working group that brings together leaders across departments, including product, engineering, legal, and sales. The group’s goal is to identify the right applications for AI, evaluate potential tools, and encourage employee experimentation through workshops, one-on-one guidance, and pilot programs.

“If everyone has to figure out AI tools on their own, we risk losing interest,” said Sarah Rich, a senior principal data scientist at CarGurus and a lead coordinator of AI Forward. “We’re trying to offer cheat sheets and share what’s working.”

She added that once employees see how AI can make their day-to-day more efficient or offer new approaches, they tend to get on board. “We want to make sure that when we ask people to invest time in AI, they’re going to quickly see a reward.”

Rich spoke with Business Insider about how AI Forward is helping employees gain the confidence to explore the technology.

The following has been edited for clarity and length.

Business Insider: What was the reason for AI Forward?

Sarah Rich: There’s a lot of pressure to get ahead with AI. And I imagine this is the case at many companies — there’s a sense that if you don’t keep up, you’re leaving innovation on the table. At the same time, there’s a gap between the excitement around AI and understanding what it means for each role.

We started AI Forward to meet every business unit and function where they are. The group works together to evaluate use cases and AI tools, which is key given how fast AI is evolving and the constant onslaught of capabilities. The group also offers structured support to help employees learn how to use the tools.

How often does the group meet, and what was your first order of business?

We meet monthly as a group, and in between, there are focused sessions within their respective departments.

One of the first things I did was meet individually with leaders to help identify a few solid use cases that could really move the needle for their teams. Some were ready to go; others had no idea where to start. We spent a lot of time brainstorming, understanding where the underlying tech is, and recognizing that in some functions, the tech just isn’t there yet.

But in other functions, like coding tools in engineering or natural language-based solutions for reviewing contracts in legal, the tools are ready.

What happens next?

We carve out time and space for people to experiment. For our engineering teams, we run office hours and jam sessions, which are essentially open collaborations, to help people learn coding tools, like Cursor and Windsurf. We also held an AI coding week to help everyone start using an AI tool on the job.

LLM solutions are effective for language-focused work that’s labor intensive. When teams experiment with those tools, they see their work accelerate quickly. We make time for experimentation; it doesn’t just happen. But usually people see something that impresses them, and AI starts to sell itself.

What’s the group doing to support employees who are less open to AI?

People are at different places on the adoption and enthusiasm curve. Some are excited about an open-ended jam session. Others need structure, where they’re required to try a tool on ticketed work, or assigned tasks or projects, and get help as they go.

Our group has learned that we need offerings at different levels. It’s important that everyone comes along to some degree, but not everyone is going to have the same level of zeal, and that’s OK.

How are you measuring success for AI Forward?

We’re tracking several metrics: how often people use AI, which tools they use, their confidence in using them safely, and their overall sentiment about AI.

There’s often a focus on adoption in terms of efficiency or hours saved, but people tend to misjudge that. AI might not always save time, but it might help you create a better product because you explored six different directions to test options before feeling confident you’ve landed on the best one. We’re careful about sentiment because AI is disruptive and can feel threatening. Pushing AI without acknowledging that nuance feels tone deaf.

What have you learned from AI Forward?

We’ve seen patterns emerge in our data in three phases. First, people feel enthusiastic because they’ve been told AI is magic and will solve everything. Then, there’s this middle-ground disillusionment, where people have had some interaction with AI tools, but they haven’t worked or lived up to the hype. There’s a narrative around AI replacing jobs versus augmenting them.

The ideal third phase comes when people start to use AI and don’t feel threatened by it. They see that it makes them better at their job. They also get that without real people, AI can’t do meaningful, impactful work.

Sentiment depends on where the individual or team is in their adoption effort and how successful they’ve been at finding the right use cases. Based on internal data ranging from the use of enterprise-wide AI productivity tools, procurement requests for new AI products, and anecdotes across teams, it’s clear that a vast majority of employees have, at minimum, tried AI in their day-to-day work.

What’s your advice for companies that want to start similar AI working groups?

Even though AI is novel in many ways, especially in how it affects people psychologically and emotionally, it’s also pretty familiar.

While there’s a tendency to get caught up in technology, the real challenge is the humans. I recommend focusing on them: bring people together, make them feel safe, and give them a reason and a space to pay attention. It needs to feel good and encouraging, not alienating.