The CEO of a McKinsey-backed employment nonprofit shares 4 ways leaders can implement AI effectively

• The CEO of a McKinsey-backed nonprofit says too many firms roll out AI without a clear usage plan.
• Generation found earlier in 2025 that most workers use AI weekly, but half are self-taught.
• CEO Mona Mourshed urged leaders to set guardrails, empower "power users," and treat AI as a multiplier.

Mona Mourshed has spent over a decade working on the future of work.

As CEO of Generation, one of the world's largest employment nonprofits, operating in 17 countries and helping more than 140,000 people land jobs, she has a front-row seat to how companies are grappling with artificial intelligence.

Her takeaway: Many companies are rolling out AI without a clear strategy.

"The vast majority of employers are rolling out AI tools in some shape or form," Mourshed, who previously worked at McKinsey — which founded Generation — told Business Insider.

"The question is whether they are rolling them out in a way that is effective."

Too often, she said, companies take a scattershot approach.

"For many employers, it is, 'Hey, here's a license, go ahead and use it,'" she added. "As a result, the employee doesn't know how or why you are supposed to use it so that you get the much-wanted gains of productivity, quality, and satisfaction."

Generation's own research highlights that gap.

In a survey of over 5,000 people across 17 countries in early 2025, the nonprofit found that 65% of respondents were alread. y using AI on the job.

Nearly 80% of those used it at least weekly, but 52% said they were self-taught, relying on tutorials or colleagues rather than formal employer guidance.

Here are Mourshed's four recommendations for CEOs who want to adopt AI effectively.

Start with the use case, not the shiny tool

Mourshed said the biggest mistake is giving employees AI tools without linking them to a problem.

"The difference is not identifying the use cases," she said.

She cited an example from Generation itself.

"We are an employment organization. There are certain things that are bottlenecks to our growth. We need to mobilize a lot of jobs," she said.

"So that's the first question. Our big bottleneck to growth is mobilizing jobs. So, how can we use AI to help us to mobilize more jobs faster? That then begins the conversation."

Leaders who start with a bottleneck, rather than a tool, are far more likely to see measurable gains, she said.

Build clear guardrails

AI can only work if leaders are precise about workflows and careful with data.

"AI is a tool. It needs to be fed very detailed workflow steps, and it needs to be fed data in order for it to do its thing," Mourshed said.

She warned leaders to balance access with responsibility.

"You have to feed it data, but you also have to be very careful that you're not feeding it personal information," she said. "You want to make sure that we are mitigating bias, so you, in some cases, don't want to provide data about gender or about ethnic backgrounds."

Empower internal champions and safe spaces

Change doesn't just come from the C-suite. According to Mourshed, companies need to identify and elevate "power users."

"There will be people who take to it and start using it all the time," she said. "These people are actually the best sherpas for everyone else."

At Generation, she encourages staff to share their experiences in AI "roundtables."

"Think of it as the water-cooler coffee chat — we do it virtually, but it's around AI topics," she said.

Treat AI as a talent multiplier, not a replacement

Some sectors are already seeing a drop in entry-level roles exposed to AI.

"There is a reality today that when you look at entry-level vacancies in professions that are AI-exposed, entry-level vacancies in professions that are AI exposed, we do see a decline. And that's not just in high-income countries, it's in middle-income countries as well," she said.

But she cautioned against a doomsday narrative.

"We also know from the history of technology that as things change, technology also gives rise to other jobs," she said.

Even in industries that may seem insulated, she said, AI is quietly reshaping workflows.

"We are seeing that AI is coming into play in a quality-assurance role," she said, pointing to examples like solar panel maintenance and garment inspection.

The bottom line: Don't sit back

Mourshed said leaders can't afford to sit on the sidelines while AI evolves.

"This is a period of tremendous learning," she said. "There is no magic bullet here. We all have to slog through it in order to understand how to make the most of this technology."

She advises CEOs to look both outside and inside their organizations: learn from peers, track internal champions, and spread their practices.

That, she said, is what will separate the winners from those who fall behind.