As AI becomes increasingly prevalent in our personal lives, CIOs can expect rising demand to adopt similar technologies in the workplace. To bring AI on board effectively, however, CIOs will need to balance a fast and comprehensive roll-out of these technologies with the security and privacy issues that AI adoption inevitably raises.

First, CIOs should start by mapping the current state of the organization and the vision for why and how the proposed initiative can be used by your organization. Then move with a plan of action with clear quantifiable success criteria with smaller pilots of AI to “test the waters” before initiating larger-scale deployments. Such pilots not only allow CIOs and their teams to iron out potential flaws or vulnerabilities with minimal risk to the enterprise. They also give line-of-business employees the chance to get used to working alongside AI assistants, whether relying on them for decision-making support or using them to streamline collaboration with other individuals and teams. The results of these pilots can help CIOs design both infrastructure and processes which maximize the AI’s utility across the organization.

CIOs must also monitor the effectiveness of AI in the workforce on an ongoing basis. AI’s greatest strength comes from its ability to adapt to feedback from its users, to deliver better results over time. The wise CIO will constantly canvass opinion from those using the AI on how accurate, efficient, or compliant its performance is, and take steps to adjust its wherever necessary.

Finally, CIOs will need to work with the rest of the C-suite to set goals, measure progress and upskill their employees for a world where working with AI-based “co-bots” (collaborative robots) is no longer the exception but the norm. That will involve not only technical training in some instances, but also ensuring that employees are flexible and open to constant learning. AI’s role in workplace collaboration will continue to grow, but its exact trajectory is still impossible to predict. The more comfortable employees are with change, the more effectively they’ll be able to work with AI’s evolving capabilities–and one another –to create value in ways that only humans can.