A step-by-step guide to implementing AI at work

Artificial intelligence continues to prove itself a valuable asset to business, but a company's path toward more AI innovation can seem complicated at first. Where — and how — do you start?

Seventy-eight percent of C-suite leaders say their organization is using AI today, according to a recent report from software company UKG, and another 71% of leaders say advancing their use of AI is a priority for their organization in an effort to stay competitive. But many are still struggling to figure out where to begin. 

"AI has been around for a long time, but generative AI capabilities are really changing everything," says Vinay Gidwaney, chief product officer at insurance, financial services and HR Consulting platform OneDigital. For his own team, embracing AI starts with education, he says. "I recently did a presentation to our entire growth marketing team about the role that AI will have in their daily work — from helping generate copy to integrating data more effectively." 

Read more: Are your company's AI decisions good for employees? 68% of C-suite leaders aren't sure

Companies across every industry have been experimenting with AI best practices. According to a 2022 study by IBM, 26% of organizations are using AI for marketing and sales, 52% are using it to improve IT, and 48% are seeing a better customer experience. Tools that are diverse in performance, however, can breed confusion and trepidation.

"We have to tease apart what we mean by AI," Gidwaney says. "Guidance is knowledge. All of a sudden, we have this powerful tool to help us generate knowledge and generate advice at a higher scale, at a different level of quality and detail." 

Gidwaney recently chatted with EBN about the strategies he and his company have used to ease themselves, and their employees, into generative AI. 

What was OneDigital's approach to implementing AI?
We think about AI adoption in three different ways. One was embracing the adoption of generative AI in every software product we use today. Frankly, employers don't have much of a choice when every single software vendor out there is including generative AI as part of their software offering. We leaned in and  decided that we were going to train our people on how to use this technology, so that they knew how to do it safely instead of resisting it. 

The second strategy is embracing Microsoft as a software partner in deploying generative AI functionality across the entire organization. Microsoft has done an amazing job innovating in this space, so we're going to start rolling out Microsoft tools across the entire organization. People are using ChatGPT anyways, so I'd rather that they do it in a place where our information and our data is protected. 

Read more: Why isn't the C-suite listening to their workers?

The third area is really focused on this idea of AI as a companion. When serving clients, many of our teams are dealing with information overload — too many meetings and too much to keep track of. If you can use AI to help you summarize meetings, make sure you get action items out, help you address clients needs, that's super effective. 

How did you manage your own skepticism around AI? 
We've always had a very innovative culture here. I've been at the company for more than two years and one of the first things that I talked about when I joined this organization is that we have to always embrace new ideas. Thinking about your culture is really important — what are the cues that you're sending and receiving from your people about taking risks and thinking about new ideas and new ways to work? We also need to have an open door policy to this: I have office hours where folks can come in and ask questions and learn what the latest thing is.

How did you make employees a part of the conversations you were having about AI? 
One of things we've done is centered our employees in every conversation — it's always been about supercharging them, not replacing them. It's about taking something they do today and allowing them to do it better. The second thing we do is we ask for feedback. When we decided to lean in on AI, I stood up in front of everybody through multiple forums and said, "Listen, we're going to learn how to do this and we need you to tell us what you're doing, how are you using these tools and what difference it's making." It's not a one way street, it's a bi-directional feedback loop.

Read more: Worried AI will steal your job? New data shows it could increase headcounts by 9%

What are some safeguards you've put in place since rolling out more AI tools? 
Training. We found partners that helped educate us on how to do this and train our leadership team on what AI was all about. Second, you have to have a policy. People need to know what's okay and what's not okay. Using tools on the internet and exposing data is a tricky issue and there's copyright issues and privacy issues to worry about that. So we made it very clear upfront that there's a policy, which will evolve as we learn more. We've also been talking to our partners and our clients about what they need from us and having conversations with them, so that they understand what we're doing. It's a multifaceted approach.