Catgory: companies, world-ai
Almost every CHRO surveyed says AI is central to their company's strategy. Half of them don't think their managers are ready to lead employees through it.
That gap comes from Gallup's Global CHRO Roundtable, which surveyed 102 chief human resources officers. Ninety-nine percent said AI is somewhat or very important to their organization's strategy. Fifty percent said they're not very confident, or not at all confident, in their managers' ability to guide employees on using it.
Culture Doesn't Automatically Get Better
Separate Gallup workforce data suggests that concern is well founded. Employees at companies that haven't adopted AI mostly report their culture has stayed the same over the past year, split evenly between the small share who say it's improved and the small share who say it's worsened.
At companies that have adopted AI, fewer employees say things stayed the same. But the shift isn't consistently positive. About a quarter say their culture has worsened, and almost as many say it's improved.
Adopting AI, in other words, doesn't reliably make a workplace better or worse. It just makes change more likely, in either direction.
The Manager Effect Is Enormous
One factor separates the companies where AI is helping from the ones where it isn't, and it isn't the technology itself.
Employees who strongly agree their manager champions AI are dramatically more likely to say AI has actually transformed how work gets done, 33% compared with just 4% among employees who don't feel that support.
The culture numbers follow the same pattern. Employees with actively supportive managers are considerably more likely to say their culture improved over the past year, and nearly three times as likely to say it improved a lot.
Same technology. Same rollout timeline, in many cases. Wildly different outcomes, depending entirely on the manager in the middle of it.
That's a harder problem for most companies to solve than picking the right AI tool. Software rollouts have a defined start and end date. Building a manager's confidence and skill in guiding a team through change happens gradually, and it's much easier to underinvest in because the payoff isn't immediately visible on a dashboard the way tool adoption metrics are.
Training Managers Isn't the Whole Fix
CHROs clearly see this. Most surveyed said they're encouraging peer learning, and 57% said they're now providing AI training specifically for people managers. A majority are also building centers of excellence or designating internal AI champions.
Those are reasonable responses. They're also probably not sufficient on their own.
Gallup's broader research has consistently found that managers already carry expanding responsibilities and wider spans of control than they used to. Handing them AI transformation on top of that, without adjusting their workload or giving them real support, risks turning a genuine priority into just another item competing for the same limited time.
There's a sequencing problem hiding in these numbers too. Many companies are training managers on AI tools at the same time those managers are being asked to model AI use for their own teams, which leaves little room for a manager to build real confidence before they're expected to project it to others.
What This Means for Miami
South Florida's economy runs heavily on industries where the manager-employee relationship is the actual product: hospitality, healthcare, financial services, real estate. These are workplaces where a poorly supported manager doesn't just slow down an AI rollout, it shows up directly in guest experience, patient care or client retention.
Miami companies investing in AI tools without a parallel investment in manager capacity are likely to see the same split outcomes Gallup found nationally, some teams thriving, others souring, largely based on how much room their managers actually had to lead the change well.
For Miami's HR and people-operations leaders, the practical takeaway isn't to slow down AI adoption. It's to ask, before rolling out new tools, whether the managers expected to champion them have the bandwidth and support to do it, not just a training session that checks a box.