More than 80% of executives say they feel some level of urgency to incorporate generative AI into their organizations, according to a new survey from Slack’s Workforce Lab. And around 50% of leaders say they feel a high degree of urgency.

However, more than 40% of respondents to Slack’s survey of 10,000 desk workers say they’ve received no guidance on how to integrate AI into their work. That keeps them from even experimenting with AI, setting them up for major productivity problems, and ultimately forcing organizations to fall behind their competitors. 

“The vast majority of people who are using AI and automation are already starting to experience productivity gains,” says Christina Janzer, head of Slack’s Workforce Lab, in a statement. “But the data indicates that failing to provide guidance or instruction on AI may be inhibiting your employees from giving it a try. If you’re looking to ready your workforce for the AI revolution, you can start by providing guidelines for how AI can be used at work.”

For the workers who say their companies have provided guidance on implementing AI into their day-to-day, 80% say they are already more productive. Desk workers spend just over 40% of their time at work on mundane, “low-value” tasks, according to Slack’s findings. Those implementing AI tools are using it to help with writing, automating workflows, and summarizing content.

But even with those productivity gains, workers appear to be split when it comes to accepting gen AI and automation into their day-to-day life. Roughly 42% of desk workers say they’re excited about AI handling tasks in their current job, 27% say they’re concerned, and 31% take a more neutral stance, saying they’re in a wait-and-see mode. 

Business leaders, however, don’t have the luxury of sitting on the sidelines to see how everything pans out. Workplace adoption of AI jumped 24% compared with last quarter, according to Slack, and there were more than 30,000 mentions of AI on company earnings calls from January 2022 to September 2023. The pace of change is moving faster than ever, and continued adoption of AI and future technologies as companies push to get leaner while improving efficiencies is only going to continue. David Crawford, global head of the technology practice at consulting firm Bain & Co., said in a statement last year that “companies that take a wait-and-see approach in terms of AI are at risk of being left behind.”

Of course business leaders moving more slowly on AI implementation have their reasons. Slack found that some of the top concerns from executives when it comes to building AI into their operations were data security and privacy (44%), AI reliability and accuracy (36%), lack of expertise and employee skill gap (25%), ethical and compliance issues (17%), customer trust and acceptance (17%), and cost of implementation and maintenance (16%).

From Janzer’s point of view, there’s room to compromise. It’s good, she says, that some executives are keeping data security, privacy, and reliable AI top of mind. But working those things out within an organization has to come side by side with guidance for workers so individual work functions can begin experimenting with AI.

“There’s a big disconnect between this executive pressure and where workers are. You have to keep those two things in parallel,” Janzer tells Fortune. “AI doesn’t need to be this separate thing we’re figuring out over here. You want it to be where the work is happening.”