ServiceNow CEO Bill McDermott recently struck a nerve across many organizations by saying that graduate unemployment “could easily go into the mid-30s in the next couple of years”.
On the surface, there is logic behind the claim. AI is indeed being positioned as a driver of efficiency, particularly the repetitive, admin-heavy processes traditionally handed over to new starters. McDermott went on to say that “so much of the work is going to be done by agents. So it’s going to be challenging for young people to differentiate themselves in the corporate environment.”
Over the last few years, ServiceNow has invested heavily in enabling the exact agentic AI functionality that McDermott describes here, through tools like Now Assist AI Agents, the AI Control Tower, and Autonomous Workforce.
But how realistic are McDermott’s predictions? And what are the implications for enterprises, employees, and graduates if the trend is correct?
AI and Graduate Jobs: What’s Actually Happening?
Before we discuss McDermott’s comments, it’s helpful to first understand what AI-related job displacement we’ve already seen – and where.
Based on the most recent data, graduate unemployment in the US sits at 5.6%. This is slightly higher than the headline unemployment figure (i.e. all workers) of 4.2%. While both numbers have risen somewhat over the last year, graduate unemployment has risen slightly faster. Since May 2025, it has risen by 0.7 percentage points, compared with a 0.4 percentage point rise for the whole economy. While both numbers have risen, neither can be considered a huge trend.
Against this backdrop, a number of high-profile research studies have been published in recent months. These aim to understand how much recent job displacement is actually a result of AI.
For instance, a recent study by Anthropic compared job market data for AI-exposed and non-AI-exposed roles against theoretical AI capabilities and real-world usage data. The report found “no systematic increase in unemployment for highly exposed workers since late 2022, though we find suggestive evidence that hiring of younger workers has slowed in exposed occupations”.
Elsewhere, a similar report by Goldman Sachs posed similar questions. It predicted that, at most, 2.5% of employment is currently at risk of automation – and that total unemployment would increase by half a percentage point during the transition. The report went on to explain that “while these trends could broaden as adoption increases, we remain skeptical that AI will lead to large employment reductions over the next decade.”
On the surface, this would suggest that concerns around AI-related job displacement are overblown. But both Anthropic and Goldman Sachs agreed that the changes we have seen are disproportionately affecting graduates and younger workers. Both also note that there remains significant untapped potential around AI adoption – suggesting the unemployment figures could yet rise further.
So, does McDermott have a point?
The Graduate Job Market Is Changing
Steven Dickens, CEO and Chief Analyst at Hyperframe Research, suggests that McDermott’s comments are a “likely worst-case scenario, meant to provoke action, rather than a statistical certainty… The most recent graduate unemployment rate is around 5%. For it to jump to 30%, we would need a total systemic collapse of entry-level hiring”.
Nonetheless, Dickens agrees that employment for graduate workers is under pressure. “Recent studies indicate that employment for workers aged 22-25 in AI-exposed roles has dipped… This is a worrying trend,” he says. This is partly related to what he describes as the ‘Junior Gap’: “Companies are finding it so easy to use AI for ‘first drafts’ that they are skipping the hire of a junior to write them… The apprenticeship model is being disrupted in real time.”
This correlates with the data we described above. Graduate roles are under pressure, and there is certainly some evidence of a slowdown in hiring activity here. But on current data, the 30% graduate unemployment figure seems like an outside possibility, rather than a continuation of the trend we’re on. In short, something fundamental would have to change for this prediction to be correct.
However, Dickens also suggests that unemployment might not be the only useful measure of AI displacement. Instead, he says: “a more realistic concern is underemployment, where graduates take roles that don’t require their degree. The risk isn’t necessarily that graduates won’t have jobs, but that the ‘traditional’ entry-level, white-collar path is being redesigned.”
So while McDermott’s comments may seem like a worst-case scenario, it’s clear that something is changing in the graduate employment market.
Why Is AI Coming for (Some) Graduate Jobs?
So why is AI disproportionately affecting graduate workers?
“Software company messaging is targeting entry-level work,” said Jessica Davis, principal analyst at Omdia. “That’s where it’s easier to make the initial ROI case. Automating repeatable, lower-complexity tasks reduces headcount costs, and vendors are absolutely selling that story to enterprise buyers right now.”
We can see this phenomenon at play already across the ServiceNow ecosystem. Here, AI agents are already taking on ticket resolution, workflow routing, and administrative tasks. However, analysts and practitioners say the outcome is not predetermined. Instead, it depends on how organizations choose to deploy AI.
John Asquith, ServiceNow lead at consultancy partner Klarus, warns that focusing narrowly on efficiency could create exactly the scenario McDermott describes:
“AI agents are already taking on the routine work that has historically been where early-career talent learns and grows. If that trajectory continues, graduate unemployment at that scale is not an exaggeration – it is a foreseeable consequence of short-term thinking.”
“When AI is introduced primarily to reduce headcount, organizations create unemployment in the short term and a knowledge vacuum in the long term.”
So, removing entry-level roles may improve margins today, but it raises a fundamental question: Where will the next generation of platform specialists come from?
A Shift Up the Value Chain?
Calitii, a Synechron company, launched its inaugural graduate training program for developers and business analysts in 2025. Director of Academy, Bushra Akhtar, argues that the impact of AI depends on how organizations define graduate roles in the first place.
“The view that AI will reduce graduate employment only applies if you’re hiring graduates for low-value, repetitive work. If you’re training graduates to make a real business impact, they won’t be replaced by AI – they will be using, building, and managing it instead.”
Arguably, as ServiceNow embeds AI deeper into its platform through copilots, automation, and increasingly autonomous agents – demand for talent moves up the value chain.
Therefore, if AI removes much of the repetitive work that once acted as a training ground, organizations will need to rethink how early-career professionals build experience. Asquith argues that responsibility lies with employers to redesign roles accordingly:
“What AI cannot replicate is judgment, context, and the ability to apply real expertise to complex situations. Those are the capabilities young professionals need to develop – and employers have a responsibility to create the conditions where that growth can happen.”
That could mean earlier exposure to complex problem-solving, closer collaboration with experienced professionals, and working alongside AI agents from day one.
The Future of Graduate Employment
While there are many different views on the future of the graduate job market, it’s clear that AI has had some impact so far – and will likely continue to do so in the future. And while McDermott’s predictions could seem to some like an exaggeration, that doesn’t mean we can ignore the situation entirely.
While nobody can predict the future, the current situation clearly points to a tightening in the graduate job market. At this point, it remains anybody’s guess how far that trend will go and what the ultimate consequences will be. In the coming months and years, this will become an increasingly vital concern for employers and wider organizations.