Senator Mark Warner: Nobody’s Ready for What AI Could Do To Us
TL;DR
Warner says flatly that neither government nor society is ready for AI’s next 3-5 years — he told Alex Kantrowitz that the short-term economic disruption could be “jaw-dropping,” with Washington still operating linearly while AI may be moving exponentially.
He’s hearing private evidence of white-collar hiring freezes already tied to AI — Warner cited big-name firms cutting interns and first-year hires, including a nationally known law firm pausing first-year associate hiring, and a midsize business shrinking a 23-person back-office team to 3.
His biggest concrete fear is recent college grads getting crushed first — with recent graduate unemployment already around 9%, Warner predicted it could hit 30% as entry-level business, analyst, coding, and other white-collar roles disappear before policymakers even start measuring the damage.
Warner’s bipartisan AI agenda is intentionally modest because Congress still lacks basic data — he mentioned bills with Josh Hawley to make the Bureau of Labor Statistics track AI-driven job loss, plus legislation on AI’s financial-market impact and a broader “economy of the future” commission.
He called the Pentagon-Anthropic clash a dangerous precedent for every U.S. tech company — Warner argued that if Defense Secretary Pete Hegseth can label Anthropic a “supply chain risk,” it becomes a potential political “death warrant” not just for Anthropic but for OpenAI, Google, Amazon, and others.
The public backlash may hit AI through data centers before it hits through model regulation — Warner pointed to ugly polling and Virginia’s fight over extracting $500 million to $1 billion a year from the industry, urging companies to proactively fund local benefits and worker transition instead of repeating the social-media playbook.
The Breakdown
“I don’t think government’s ready. I don’t think society’s ready.”
Alex opens by saying he’s been “freaked out” after hearing from AI lab insiders that progress may actually be exponential, not just marketing hype. Warner doesn’t soften it: he says the AI optimists themselves seem to be downplaying things now because they’re scared of scaring everyone else, and he thinks the next 3-5 years will bring economic disruption the country is not prepared for.
The private signals: no interns, no associates, far fewer entry-level jobs
Warner says what’s driving his alarm isn’t just lab rhetoric but what he’s hearing from companies behind closed doors. He mentions firms halving intern and first-year hiring, a nationally known law firm pausing first-year associate hiring altogether, and one midsize company bragging that AI let it cut a 23-person back-office function down to 3 — “isn’t that amazing?”
The data gap, and the fear that new grads are first in line
His main complaint is brutally simple: the U.S. isn’t even properly measuring AI job displacement yet. That’s why he’s pushing a bipartisan bill with Josh Hawley to make the Bureau of Labor Statistics track not just jobs lost, but jobs that would have been created — because he thinks recent college grad unemployment could go from 9% to 30%, especially in fields like business administration and entry-level analyst work.
Why Congress still feels miles behind the curve
Asked how many senators even know what Claude is, Warner laughs but basically concedes the broader problem: most members don’t understand this stuff, and when people feel lost, they tend to punt. He says his own AI bills — on labor data, financial-market risk, and an “economy of the future” commission — are thoughtful but incremental, even as the real moment may demand something much bigger.
Social media was bad; AI is the same failure mode, only much faster
Warner draws a direct line to Washington’s inability to regulate social media, noting that the companies all welcomed “meaningful regulation” until lawmakers put actual words on the page. But he says AI is vastly bigger: deepfakes, chatbot relationships that now show up statistically, stories of AI linked to suicide, and a labor shock that could hit within 2-3 years instead of unfolding over a decade.
The public may revolt through data centers before it revolts through AI policy
When Alex brings up ugly polling on data centers, Warner says that’s where anti-AI politics may first become tangible. He points to Virginia — the data-center capital of the U.S. — where policymakers are debating extracting $500 million to $1 billion annually from the industry, and says companies should stop fighting every demand and instead fund affordable housing, power protections, and broader economic-transition support.
Anthropic, Hegseth, and the “death sentence” precedent
In the second half, Warner zeroes in on the Pentagon conflict with Anthropic and says the real issue is bigger than one company. If Pete Hegseth can unilaterally declare Anthropic a “supply chain risk,” Warner argues, that becomes a precedent for politically blacklisting any major U.S. tech company from federal and downstream commercial business — a power he calls a “big freaking deal.”
AI warfare, surveillance, and the spooky future nobody has scoped out
Warner says he still doesn’t know whether the administration is building AI-based surveillance of Americans, and he’s frustrated that oversight has been weak. He’s open to debate on defensive autonomous systems, but insists the country should not sleepwalk into AI weapons without a human in the loop, ICE or DHS using AI for targeting, or foreign deepfakes wrecking the 2026 election before Congress even decides what the rules are.
A last note on bipartisan allies, ethics, and why he still sounds alarmed
Near the end, Warner name-checks Republican Senator Mike Rounds as someone ahead of many peers on AI and weapons, reinforcing his “don’t make this partisan” theme. But his overall mood never really lifts: whether he’s talking about Congress stock trading, crypto ethics, or AI itself, the recurring point is that institutions are moving too slowly while the technology is moving very, very fast.