According to CNBC, Democratic Senator Mark Warner of Virginia warned at the CNBC CFO Council Summit that unemployment among recent college graduates could reach a staggering 25% over the next three to five years if AI job displacement isn’t proactively addressed. He noted current grad unemployment is already at 9% as of November. Warner, who is working with Republican Senator Josh Hawley, fears the bigger issue is “jobs not being created in the first place,” citing that major banks have already cut interns and first-year hires in half. He stressed that without collaboration between government and industry to collect data on AI-linked losses, the U.S. risks a wave of “anti-innovation populism.” Warner, a former tech investor, admitted past solutions like pushing everyone to learn coding were wrong.
The real fear is phantom jobs
Here’s the thing that makes Warner’s warning different. It’s not just about the layoffs we see in headlines. His “biggest fear” is more insidious: the jobs that will simply never exist. When a big bank uses AI to handle tasks that used to go to a fresh analyst or intern, that entry-level position vanishes from the universe. It’s a silent contraction of the job market. And that’s a disaster for a system built on the promise that a college degree is a ticket to stable employment. We’re talking about a generation already saddled with debt walking into a workforce that’s shrinking at the entry point. No wonder he says “We will have a lot of parents pissed off.”
The messy hunt for data
Warner’s plan, basically, is to start by measuring the problem. He wants to work with companies—not just tech, but banks, accounting firms, chip makers—to track AI’s impact on hiring. But he concedes this is incredibly hard. Who exactly do you involve? How do you define an “AI-linked” job loss versus a regular business downturn? The government’s track record here isn’t inspiring, as he himself admits by pointing to the failed social media guardrail efforts of 2017-2018. So we’re supposed to believe this time will be different? The alternative, he argues, is a political backlash so severe it could stifle innovation itself. It’s a compelling argument, but getting a bunch of competitive corporations to hand over sensitive hiring data sounds like a monumental task.
We’ve already gotten the solutions wrong
I think the most refreshing part of Warner’s comments is his blunt admission that the previous playbook was a failure. “The answer was to make everyone a coder, and that was not the right solution set,” he said. We’re seeing the same cycle now with vague promises about “AI-centric jobs” or a resurgence in trades. But as Warner notes, when’s the last time a CEO at that summit encouraged their own kid to be a carpenter? There’s a massive gap between the theoretical jobs of the future and the actual, dignified career paths we’re preparing people for. Throwing around ideas like universal basic income, which Warner clearly dislikes, feels like giving up on the core problem of matching people with purposeful work.
A repeat of social media failures?
The shadow hanging over this whole conversation is social media. Warner uses it as a direct cautionary tale: lots of hearings, lots of concern, and ultimately, “We did none of it.” Now he’s betting his re-election in 2026 on this being the defining issue. That tells you how serious he thinks it is. The potential disruption from AI makes social media’s impact look “tiny” in comparison. So the pressure is on. Can a divided government and a profit-driven industry actually come together for a long-term, structural solution? Or will we look back in five years at a 25% unemployment rate for grads and wonder why we just held more hearings? The clock is ticking, and so far, we’re not even sure how to set the timer.
