Join/Renew
The career services profession is navigating a fundamental tension: The labor market has changed faster than the advising frameworks in which most practitioners trained. Lightcast's 2026 Fault Lines report indicates that a third of skills in the average U.S. job changed between 2021 and 2024 — yet the majority of career advising still operates on credential-to-career logic designed for a stable market that no longer exists. Meanwhile, a
2025 NACE survey of 448 institutions found that only 40% of career centers have provided any AI training for their own staff, and a 2025 Handshake survey found that
62% of graduating seniors are worried about AI's impact on their careers. Students know something has shifted. The question is whether our advising practice has shifted with it.
This article identifies three specific, actionable shifts that career services professionals can make now — not by abandoning what works but by updating the assumptions that no longer hold. These shifts do not require new tools, additional budget,
or institutional approval; instead, they require a change in framing: what we actually help students build, how we talk about the labor market, and what role AI plays in an advising conversation. Each shift is grounded in current labor market research
and immediately applicable in practice.
The Three Shifts
#1 — From Job Titles to Skill Narratives
Old assumption: A student's degree aligns with a set of jobs, and the advisor's job is to help them find and apply for those positions.
New Reality: This assumption has already started to fade in recent decades, but a new reality requires a more dramatic shift. Only 6% of workers in the fastest-growing occupational category in the U.S. economy — AI roles — have AI-related academic degrees, according to Lightcast's 2026 Fault Lines report. Workers are entering careers through skills and adaptability, not credential matching. In fact, the student in front of you may end up working in a role that does not exist.
Recommendation: Move from organizing advising conversations around job titles toward organizing around skill clusters — specifically, which of the student's developing skills are durable across AI disruption and which are vulnerable. A student who leaves your office understanding their transferable capability set is better equipped than one who leaves with a list of job titles to search on Handshake.
#2 — From Career Plans to Adaptive Strategies
Old assumption: A five-year career plan gives students direction and confidence.
New reality: Deloitte's 2025 Future of Work research establishes that the half-life of technology skills is as short as 2.5 years. The Lightcast data show skill requirements changing year over year in nearly every sector. A five-year career plan built on specific job titles and skill requirements is likely to be misaligned with the actual labor market before the student reaches the halfway point.
Recommendation: Five-year plans should be replaced by a quarterly career strategy review habit. Help students build the metacognitive practice of checking in on labor market signals, reassessing their skill positioning,
and adjusting their strategy. The goal is not a fixed plan — it is an adaptive orientation toward uncertainty. Students who graduate knowing how to update their strategy are more durable than students who graduate with a polished plan that can
quickly become outdated.
#3 — From AI Tool Hand-off to AI Literacy Coaching
Old assumption: Pointing a student toward ChatGPT or an AI resume tool is sufficient AI integration in career advising.
New reality: Stanford's February 2026 AI+Education Summit research found that between 70–80% of students use AI to short-circuit learning rather than enhance it. A student who produces an AI-generated resume without understanding why each word is there has not developed career readiness — they have outsourced it. Worse, Stanford researchers found that students who rely heavily on AI for creative tasks begin believing AI is more creative than they are. In career development terms, this is an identity problem, not a tool problem.
Recommendation: Treat AI literacy as a coaching topic, not a tool recommendation. Help students understand what AI can and cannot know about them — it can process labor market data, but it cannot know their values, their risk tolerance, their family context, or the difference between what they say they want and what they actually mean. An advising conversation that builds a student's ability to use AI critically — as an intelligence layer they interrogate rather than an answer machine they defer to — is qualitatively different from one that ends with "Go try this tool." That difference is what the students who most need us — first-generation, low-income, and underrepresented students navigating a system that was not built for them — actually require.
These three shifts share a common logic: The advisor's irreplaceable value is not information delivery — AI does that faster. It is judgment, relationship, and the human capacity to help a student understand who they are becoming, not just where they are applying. That value has never been more necessary. The question is whether our practice reflects it.
