AI is reshaping entry-level work before they start

AI is – As the class of 2026 steps into white-collar jobs, managers are already changing what “junior work” looks like—handing new hires faster, larger responsibilities through AI tools like ChatGPT. But the advice from leaders is consistent: use approved systems, don

For the class of 2026, graduation season may still look traditional. The first job, though, is starting to feel different. In the workplace. AI tools like ChatGPT are taking on tasks that used to be reserved for the early. “learn-the-ropes” phase—pushing some companies to give entry-level hires bigger responsibilities from day one.

Peter Cappelli, a management professor and director of the Center for Human Resources at the University of Pennsylvania’s Wharton School, described the shift in blunt terms: “AI is changing the entry-level experience for an entire generation of white-collar workers.”

With that reality moving from campuses into offices, career leaders are pushing a clear message for newcomers: the smartest path isn’t to hide behind AI or rely on it blindly. It’s to use it with judgment—while building the habits that get you trusted, promoted, and kept.

Chris Lyon. head of engineering at Twilio. told newcomers to avoid so-called “shadow AI”—the unapproved tools employees might use on their own. He framed the issue as both practical and sensitive: companies have safety protocols for a reason. and sensitive company information can end up outside corporate guardrails. “You have to be extra careful,” he said, recommending workers stick to approved AI tools.

Jeff LeBlanc. a management lecturer at Bentley University. warned against going too far the other direction—overusing AI for every part of the job. If “every email. every idea. every decision gets filtered through AI. ” he said. people risk never fully developing their own instincts. His point landed on a human timeline: early career growth is supposed to include uncertainty. mistakes. and figuring things out—because that discomfort is where confidence can be built.

Even when AI is used, output still needs a reality check. Matthew Bidwell, a management professor at Wharton, described AI as “a flaky coworker” that can be brilliant one moment and catastrophically wrong the next. His advice was direct: “You want to make sure that the output passes the smell test.”

That “smell test” isn’t just about quality control—it’s also about workplace relationships and resentment. Dr. Andrea Derler. head researcher at Visier. a workforce intelligence company. said senior employees who started their careers doing grunt work often don’t want to clean up careless AI-generated content from junior staff.

The same principle applies to asking for help. AI can make it feel easier to work alone behind a screen. but Hebba Youssef. chief people officer of the media company Workweek and host of the HR podcast “I Hate It Here. ” argued newcomers shouldn’t treat human questions as a weakness. Asking why a task matters and what success looks like is “actually a wise move because it demonstrates critical thinking.”.

Rebecca Port, chief people officer at Okta, an identity-management company, put another boundary around the chatbot fantasy. Even if you don’t need an assist, building relationships with more experienced colleagues is still smart. A chatbot “isn’t going to help you navigate office politics. ” and while performance matters. relationships can also play a part in whether you get a promotion or bonus. “Like it or not, organizations are social organisms,” she said.

Reliability, meanwhile, is the expectation AI can’t replace. Dr. Derler warned that the workplace isn’t as forgiving as the classroom. If newcomers consistently show up late, dress inappropriately, miss deadlines, or commit other faux pas, it can dent their career. She described reliability as a “hidden expectation” that’s essential for earning an employer’s trust.

Her instruction was practical and unforgiving: “If you say something will be done by 5 o’clock tomorrow morning, have it done by 5 o’clock by tomorrow morning.” She added, “Do not try to negotiate on timelines two minutes before it’s due.”

The shift underway—bigger responsibilities enabled by AI—doesn’t change the fundamentals. It changes the speed of work, and it raises the stakes for judgment. Used well, AI can accelerate what a new hire does. Used carelessly. it can land them in the same bind every junior employee tries to avoid: being blamed for problems they didn’t personally create. or being distrusted for shortcuts they took.

So the message for first jobs isn’t anti-AI. It’s more precise than that. Use approved tools instead of “shadow AI.” Don’t let AI replace your thinking. Check the output. Ask questions that reveal how you think. Stay dependable on the basics. And build relationships—because the parts of work that shape your future still live with people, not prompts.

AI in the workplace entry-level jobs ChatGPT shadow AI Twilio Wharton Twilio engineering Bentley University Visier Workweek Okta workforce intelligence career advice reliability at work