Not so many years ago, leaving college felt electric. I remember it as a moment charged with promise, rich with possibility, and alive with the feeling that a dozen doors had swung open at once. Graduates’ ambitions or enthusiasm haven’t changed today. But the job market they’re entering is unrecognizable.
Tasks that once defined a junior’s role—analyzing data, drafting reports, and taking meeting notes—are being automated, changing the shape and availability of early-career work.
The Pew Research Center reports that 73% of people under 30 now believe artificial intelligence will lead to fewer jobs, up from 61% two years ago. Their anxiety isn’t misplaced; entry-level vacancies are shrinking.
For those who do make it through the door, the nature of the work is shifting, too. The “seniorization” of junior roles means that in AI-exposed fields, junior hires are now seven times as likely to require midcareer capabilities as hires into roles less connected with AI. And rather than “learning by doing,” many are “learning by reviewing,” spending time tinkering with AI-generated outputs, rather than developing hard-won skills like judgment, pattern recognition, and critical thinking.
Against this backdrop, a problem emerges. Working effectively with AI demands a degree of experience many juniors might lack, new to the workforce as they are. Experience is what helps us distinguish helpful AI outputs from bad. While the risks aren’t solely associated with those in entry-level jobs—senior colleagues can also be guilty of overrelying on AI to crunch through an ever-expanding list of tasks—it’s reasonable to assume that junior colleagues are the most exposed to AI-enhanced errors and authoritative slop.
That isn’t to say we should limit or fear AI use. Far from it. The right tools offer measurable impact by transforming workflows, removing administrative frictions, and boosting efficiency. In a future of work that will be shaped and defined by AI, ensuring colleagues at all levels are developing AI literacy is critical.
But the core issue—and irony—remains: Judging AI’s output effectively often requires skills we now expect juniors to exhibit, even while limiting the opportunities for them to develop them. This state of affairs risks contributing to a growing early-career soft skills gap. Leaders who are serious about closing this gap and setting graduates up for future success should focus on three things.
1. Create deliberate critical thinking opportunities
Regularly giving new hires projects that require them to weigh options, make recommendations, and get feedback is one of the surest ways to build practical judgment. One approach is to give new hires projects to own. Have them present progress live to colleagues who can probe their reasoning, ask targeted questions, and understand their approach. This shouldn’t be treated as a test, but as a means for junior colleagues to develop and defend effective reasoning. It takes work offline and off models and into the real world, while exposing juniors to opportunities for critical thinking and constructive feedback from peers.
2. Make AI use visible and low-risk
Beyond setting explicit AI guidelines, leaders must keep AI use visible and low-risk, so mistakes surface early and stay contained. In practice that might mean encouraging juniors to automate only a predefined set of simple tasks—such as summarizing an internal call or condensing a data set—while logging each step in a shared document. Or asking colleagues to flag when content is created with the support of AI, so this is clear when it moves along the chain for implementation or review. The log becomes an accountability trail, giving managers visibility over AI-generated content and the opportunity to provide feedback if juniors miss where AI outputs fall short.
3. Start training in AI fluency early—and keep it going
Robust AI fluency is about knowing when AI should and shouldn’t be used, which tools fit which tasks, and having the skills and vocabulary to use the right prompts and assess AI outputs critically. These are skills that need to be learned. This might look like sharing real organizational use cases that show where AI workflows worked and where they didn’t. Organizations must also ensure juniors aren’t overrelying on AI tools for tasks where judgment, style, and creativity are critical. Once individuals know what good looks like and how to get there, they’ll be in a better position to use AI as a tool to reach that endpoint more quickly.
Crucially, every individual at every level of an organization must also learn that AI isn’t a shortcut for thinking. Nor is it a substitute for effort. As the guardians of future talent, businesses that make this clear and empower new hires to build both technical and soft skills early in their careers will see that talent soar.
For graduates, too, my message is clear: Nurture AI fluency, be curious and active in exploring new tools. But don’t do so at the expense of challenging yourself to solve problems and form opinions independently. In an AI-enhanced future of work, it’s those who can strike this balance successfully who will become the leaders of the future.
