While AI continues to be the hot workplace topic to kick off 2025, many companies have no idea what to do with the technology in practice. Actual use cases of generative AI and case studies within a business environment are not as prevalent as one would expect. David Rowlands, Global Head of Artificial Intelligence for KMPG, and Ruth Svensson, a partner at KPMG UK who serves as the Global Head of People and HR CoE, discuss why that is — and it largely comes down to not having an AI vision.

At KPMG, "AI has been a totem for everyone to get around because it's enabled people to access and understand what the technology could do for us and for our clients," says Rowlands. While AI's value is clear for companies such as NVIDIA and Microsoft, Rowlands makes the point that we haven't seen the value of AI reflected in the numbers for banks, government, and other non-tech companies. When companies fail to see a benefit, AI adoption rates slow. "This adoption problem is because the value got stuck," says Rowlands. "Our role is to adopt those large language models, take them, harness them, and apply them in situations that benefit our clients."

What's holding people back from adopting AI? According to Rowlands and Svensson, it's largely trust, building the right technology environment, data controls, proper training, understanding the human element, and having a vision. "People are asking if they can trust the AI," says Rowlands. "Can you demonstrate that you're trustworthy to your stakeholders? Have you got the fundamental technology environment? An enormous amount of enterprises just simply don't have the core fundamentals in place from which they can start to experiment and grow in their use of AI," says Rowlands. Another barrier is data protection. In fact, 77% of executives in a study conducted by MIT Technology Review cite their regulatory, compliance, and data privacy environment as a leading barrier to rapid AI adoption.

Proper training also stymies adoption. "Businesses have been told by the providers of generative AI that it's really user-friendly, so they haven't invested in the traditional upskilling mechanisms that they would do for a system implementation," says Svensson. "When you implement a new system that supports a whole process, you would train everyone on the new steps of the process. Because generative AI operates at a task level rather than at a process level, you can't see the training gaps as easily."

Svensson continues the thread of the human element of adoption — something many companies overlook when implementing change programs. "What's interesting is that if you look at the segments of society who have embraced generative AI, it's because they care about it. People adopt things that they care about. However, the reality is that the vast majority of the population does not care about using AI in their workplace because they don't see the personal benefit to them. They struggle to understand how it's going to make their life better. For example, software engineers have adopted it much more readily. When you hear software engineers and coders talking about the generative AI solutions that they've adopted, they refer to it as their best friend. They are enthused by the way it helps them work faster. Organizations, for a number of reasons, are struggling to make it feel relevant to the mass population within their workforce."

This lack of seeing a personal benefit is directly related to the lack of having a vision for the future and the technology — and is arguably the biggest barrier to AI adoption. What, exactly, are companies going to do with AI? "A large barrier is clarity of business case," says Rowlands. "The pricing of AI is neither cheap nor clear. If people aren't using it in a way that drives a strong business case, then you can get this out of kilter. For example, if your people are using AI as an enhanced search, it's practically a thousand times more expensive than Google — and never mind the damage to the environment. Every search uses a pint of water."

Those who are adopting AI at the highest rates are doing so because of their ability to see a clear vision of how it can help them get ahead. "The interesting thing about AI is that CEOs are the ones adopting it and getting it the most because they are the ones that spend the most time thinking about the future and thinking about what it could be," says Rowlands. That point takes us back to the role of vision and value in AI adoption. "The first step to adoption is to understand where the value sits," reiterates Svensson. "They have to identify where the opportunity sits, then put energy into having a systematic adoption approach. It's not just the upskilling; it's getting people excited about the technology."

While there is a robust conversation around barriers to adoption and how fear of job displacement plays a role, the paradox is that people are still adopting AI faster than prior technological advances. "Everything I've read regarding adoption is that it is still much faster than the adoption of the internet. So as far as tech adoptions go on a macro scale, it's faster — probably because it's so user-friendly," says Svensson. Organizations, however, are moving at the speed that organizational change-makers are used to — not at the speed of the technology’s evolution. "The question is 'What can we do differently as a combination of human talent and now artificial intelligence?' And the answer can't be negative," says Rowlands. "In KPMG, we talk about uplifting human potential. What will we do as humans as we free ourselves from some of the tasks that we did before, which now can be done by AI? That question is highly complex, but I'm excited about it."

But before that can happen, businesses must address the elephant in the room. "There's a huge trust issue," Svensson cautions. Businesses that focus on the vision of what AI could mean for its people and how it will benefit them versus stoking the fear of job displacement will come out winners. It comes back to the vision of what people can do with the time that AI will unlock. Cheap companies will use that time to eliminate precious capacity and slack in the business. Smart companies will allow their people to use that time more creatively and come out ahead.