Employee AI agent adoption: Maximizing gains while navigating challenges

Although Agentic AI definitely marks a turning point in the interaction between people and computer, from tool use to cooperation, the next step is to integrate these agents and actually distract value. With Venturebeats Transform 2025Matthew Kropp, director and senior partner at BCG, offered a game plan for Workflow evolution, adoption of employees and organizational change.

“The companies that are at the top of this curve – what we call Future Build, those who are the most mature – see substantial results: 1.5 times more sales growth, 1.8 times higher shareholder value,” Kropp said. “There is value here, but we are early.”

Implement, reform, invent

To take advantage and create value with AI and with agents, a company must determine where to focus, with the help of an implementation, reform, framework. AI is already used in every company and will have agents within a few years. But if you give an employee a chatbot, you have not changed the way the work was done. You must reform the work and reform functions, departments and workflows by determining where human work can be automated.

“We are currently advising companies to concentrate on your three or four large rocks. If you have a large customer support organization, you have to apply AI in customer support. It has a huge impact. If you have a large technical organization, you have to use tools such as Windsurf to reform the way you do engineering, software development.”

The invention is still at the very early stage, but companies have to think about how they use AI’s ability to be creative, reason and plan. Look at services and products, and how you deal with customers: can you reinvent that using those options? “

Make -up company L’Oreal, for example, launched a virtual beauty adviser to further scale that exclusive service outside their store locations, whereby they reinvent the way they think about interacting with their customers on a scale.

Think beyond the basic use outings

It is also crucial to think about how AI changes your company. A lot of attention has been paid in recent years to replacing cost reduction by employees, but that is not great to think. AI strengthens the employees you currently have, which dramatically increases their productivity.

“This is what we see in software development,” he said. “I don’t think companies will see that companies will dismiss their software developers. We are going to see a huge explosion in the amount of capacities and functions that software companies build.”

In a BCG study conducted with Harvard, Wharton and MIT, they asked 750 knowledge workers to write a business and marketing plan, with and without a generative AI. The participants who used GPT4 performed 25% faster, completed 15% more tasks and the quality of their output was 40% better. And when they get an LLM, the lower performers in the baseline did just as well as the top performers.

“It raised everyone’s performance, which is very powerful, because in most organizations the new connection is less effective than more experienced people,” he said. “It has the possibility to increase the time to competence.”

AI can also surpass the human scale, even open new applications that were not possible before. In the medical space, for example, the results for patients have been significantly improved with preoperative and postoperative follow-up of a nurse, but the implementation of this has been the prohibition of cost the arrival of AI nurses who can adopt that task for a large patient population.

Overcoming the biggest obstacle: adoption

Although these tools are fantastic, people don’t use them. BCG followed the acceptance of Github Copilot and productivity statistics for an organization with around 10,000 software seniorisers. The top 5% engineers doubled in productivity in four months, while 60% showed no improvement because they did not take over the tool at all.

Why will people not adopt? There are three reasons. First it is possible to arise. The second, habit in slowness. The third is the threat of identity, and that is the most difficult to overcome. Developers ask: “If this AI code can write for me, who am I then? What is my value?”

“This will be the real work from the next three to five years,” said Kropp. “It ensures that people use the agents.”

Strategies that overcome the reluctance

There are a few valuable ways to overcome these challenges. Getting the right tool is of course the first step and integrate it with the way people work by explicitly training them. It is also crucial to measure and celebrate adoption for those employees who actively use the tools, so that everyone starts to see that they have to come on this bandwagon.

Another important step is to increase scarcity – that means removing resources, so that employees have to do more with less. At the same time, it is essential to re -design work processes hand in hand with those employees who are in the front line. Do not only identify difficult processes where manual work can be automated – identify the parts where people bring value.

“We minimize the toil and we maximize the joy,” said Kropp. “We are behind with a much more efficient process, a much more efficient company, a much more productive workforce and jobs that people like to work on.”