Half of all CEOs fear for their jobs if they don’t get their AI strategies right this year, and 74% say they could lose their jobs if they don’t deliver measurable AI-driven gains by the end of 2027.
You might think that would drive senior executives to use AI tools extensively, but new research from Stanford University economist Nick Bloom finds the opposite: 69% of CEOs, CFOs and senior executives are using AI at work less than an hour a week, and 28% aren’t using it at all.
Bloom, who spoke about his research at Charter’s Leading with AI Summit last week in San Francisco, is known for his rigorous data-driven studies that have made him the world’s leading researcher on remote work.
Now, he and his colleagues are applying the same methodology to AI, partnering with reputable sources like the Bank of England and the Federal Reserve to survey more than 6,000 senior executives across four countries. “This is like 15 minutes of data nerd fest,” he joked as he began his talk at the summit.
The result is one of the most comprehensive surveys to date of executives’ usage of AI, its impact on their businesses, and their predictions for the next three years of AI’s impact on productivity and jobs. (Bloom also joined a Charter Forum session in January to preview his new data on AI and work; this column is adapted from that conversation, his just-released study, and his remarks during last week’s summit.)
No impact to date
In the aggregate, Bloom’s findings show that as of late last year, executives reported AI has had no impact on employment. Despite mass job cuts like the ones Block CEO Jack Dorsey announced last week, just five percent of executives said that AI has reduced headcount, mostly by less than five percent, and four percent said AI has increased the number of employees. The net effect is a wash.
When asked at the Forum session in January about headlines suggesting AI is prompting layoffs, Bloom said he thinks macro conditions—higher interest rates, tariffs, and tech companies needing to hit quarterly earnings while increasing investments in AI technology—are the more likely culprits. He was direct: “As of yet, there has not been a big effect.”
Bloom’s study also finds there has been a very modest boost to productivity. Ten percent of executives say they have seen some gains, but most report productivity boosts of under five percent. That’s consistent with BCG data showing roughly five percent of firms get consistent returns on their investments in AI. Atlassian data finds a similar number—four percent of executives—say AI has led to meaningful efficiency and quality improvements.
For leaders who’ve been asked to justify AI spending against near-term ROI, that’s both reassuring and uncomfortable. Reassuring because they’re not alone—many executives have struggled to leverage AI for meaningful impact—and uncomfortable because the pressure to show returns is already very present.
US executives expect the biggest hit to employment
AI’s impact is already under fierce debate, especially when it comes to its effect on employment. Bloom’s data shows executives expect more significant changes are coming, and that leaders in the US are distinctly more aggressive than their global counterparts.
Globally, 26% of executives predict decreased employment over the next three years due to AI; in the US, that number jumps to 43%. After factoring in the 15% of leaders who expect employment to grow, the net prediction by US executives is that employment will fall 1.2% due to AI. (It’s worth noting, of course, that labor laws are far more restrictive in Europe and other parts of the world than in the US.)
Compared to Verizon CEO Dan Shulman, who said in Davos “it’s possible that we see 20% to 30% unemployment levels over the next two to five years,” that might feel like welcome news. Bloom dispels that notion: a 1.2 percent drop “doesn’t sound like a lot, but that is a huge number.” In a country with 170 million workers, that’s more than two million jobs. Plus, millions of people out of work globally could have a cascading effect on the economy: people out of jobs have reduced spending power, further eroding demand and increasing the odds of a recession.
Still, Bloom introduced one potential bright spot and a “huge caveat” at the summit: The study, he noted, is of existing firms, and “there’s lots of…startups and new companies coming” that might help mitigate cuts at existing firms.
On productivity, meanwhile, the predictions run higher: 59% of US executives forecast a 2.3 percent productivity increase over the next three years.
Again, that might sound small, but a doubling of productivity growth would be “absolutely a huge number,” Bloom said at the summit. There “is nothing we have seen in the last hundred years” where technology has yet had that effect.
Put another way: If the predictions in Bloom’s data are right, we may not be in for the catastrophic outcomes on jobs currently being fretted over in viral essays by tech leaders mourning the changes in their industry. At a minimum, however, we’re likely to see millions of jobs eliminated and a widening of the K-shaped economy.
The executive-employee gap
Still, with Bloom’s data showing 69% of executives use AI for less than an hour a week, one has to ask if their perspective is accurate. The average employee uses AI more than their executive does, but not by a wide margin. Workers average 1.8 hours per week, while senior executives average 1.5 hours.
Leaders who aren’t using these tools themselves are making decisions about AI adoption, investment, and workforce planning without firsthand understanding of what the technology can really do. The people most accountable for AI outcomes are often the least experienced in their use.
Employees, meanwhile, see the AI future differently than their bosses do. They predict less than half the productivity gain that executives forecast (0.9% vs 2.3%). While employees might see minimal productivity gains themselves, that doesn’t stop executives from thinking “am I just going to get rid of that employee, or reshape that entire division?” Bloom said at the summit. Regardless of whose insight is current, he said, “there is a clear misalignment here across society.”
Three things leaders can do this week
Separate the pressure from the timeline. Board urgency about ROI from AI is real, but Bloom’s data suggests the larger economic transformation is still early. Don’t let short-term pressure push you into acting prematurely by cutting heads before it’s needed. Those who remain often lose trust in leadership.
Close the gap with your teams. The divergence between executive and employee forecasts mirrors their gaps in both enthusiasm and understanding. Leaders need to create realistic scenarios for the future and enlist teams to work toward better outcomes, not generate spin.
Use AI yourself, consistently. There’s no substitute for firsthand experience with tools your organization is deploying. The most effective transformations I’ve seen involve leaders using the tools with their teams. Your credibility in leading adoption depends on it.
For more from Charter’s Leading with AI Summit in San Francisco, check out remarks from Replit chief executive officer Amjad Masad, Khan Academy founder Sal Khan, and the chief people officers of Walmart, Airbnb, Anthropic, and Box. Full coverage of our Leading with AI Summit can be found here.
For more ideas on AI, its impact on jobs, and leading transformation, see:
- Block’s layoffs are an outlier. Their influence might not be.
- What a viral essay on AI and jobs got right—and what it missed
- Ideas on “managing the messy middle” of transformation from our New York summit