AI & Jobs: Enter Populism?
Four political attractors that could draw the battle lines in the AI-jobs debate.
Recent policy commentary has come around to discuss the impending politicisation of AI job impacts. The story goes: once people lose their jobs to advanced AI agents, surely some politicians become interested in saying something about that. That’s directionally right, but vague: it warrants a deeper look at the coalition dynamics, demographic splits, and political incentives that will drive this trend. I believe they don’t imply rampant populism everywhere. Instead, they increasingly point at Democrats running a jobs-based anti-AI, anti-big-tech message, while Republicans play coalition defense on absorption, blue-collar job growth and economic headlines.
All this matters, because AI and jobs could become the biggest story in AI policy. Decisionmakers might position around AI according to their position on its labor impacts. The first question at any AI policy press conference could be on jobs. Headlines about infrastructure projects, model deployments, and risk rules would all be read through the labor-market lens. This can even reach into mainstream, big-ticket politics: as AI’s economic impact and cultural impact increase, you’ll find more people caring about its impact on their life. And savvy politicians will quickly identify its potency as a favorable wedge through the electorate.
These trends will shape the overton window for viable proposals, determine which alliances and decision-makers matter most, and entrench positions on AI across many policy areas. But it’s hard to make predictions around how that will go: The current environment is volatile enough that either party could in theory skew pro-AI or anti-AI, both either appealing to AI’s geopolitical and economic upsides or to its labor market downsides. Neither have a solid agenda in place, as I’ve written before:
No one knows how the resulting politics will go. If I knew who was going to be the Democratic nominee in 2028, or what Democratic message will perform well in the Midterms, I’d be on Polymarket instead of Substack; and of course the actual extent of perceived and actual displacement is downstream of future capabilities and diffusion speeds. But what we can do is name underrated political attractors that could steer these trends. I believe four such attractors suggest that the GOP is on the whole less likely to polarise into a populist anti-AI stance on the grounds of jobs – less likely than people assume, and less likely than Democrats.
Being Anti-AI is attractive for Democrats
Taking a generally anti-AI position on the grounds of labor disruptions is politically attractive for large parts of the Democratic Party, especially in case the party decides on an economic populist platform and candidate for 2028. That faction already trades on skepticism of large corporations, antitrust talk, and threats to worker dignity and economic security. It maps neatly onto AI.
The anti-AI position becomes more effective still because it offers a distinction even from ostensibly similarly pro-worker GOP policies, because the GOP can easily be painted as aligned with the tech corporations. The current administration has publicly aligned with tech executives and their corporations to an extent that will be difficult to believably walk back from. In other areas of economic policy, the GOP can sometimes compete with Democratic economic populism on its own merits – e.g. by alleging to be on the side of blue-collar workers instead of city elites. But the specific alignment with AI companies makes the AI version of the Democratic argument particularly attractive and difficult to defend against to the extent that the GOP can be painted as ‘having sold out’ workers to big tech.
Incumbency hamstrings GOP Anti-AI Positioning
Yes, some similar forces apply to Republicans. Like Democratic populists, many on the right have little love for coastal billionaires; and despite recent alignment, few forgot years of perceived “woke” rule over platforms. Still, an anti-AI pivot is difficult. The first reason is coalitional conflicts: keeping the ambitious alliance between the tech right and MAGA factions alive requires careful navigation around tech policy issues. Surely, an anti-AI pivot would at once blow up that conflict, alienating a lot of the tech right, and with it fragile cultural appeal to some younger urban demographics as well as sizable amounts of donor money.
The second and bigger reason is that the GOP is presiding over the increasing salience of AI and is inseparably entangled with it. This is first a general principle: If you preside over a development, you won’t win a lot from running against it: if you preside over high inflation, no one will buy your economic message; if you preside over increasing unemployment, no one will buy your jobs message; if you entered wars, no one will buy your peace message. And if you preside over unprecented technological diffusion, datacenter buildout and tech-government alliances, no one will buy your anti-AI message.
It’s made worse by the GOP’s direct involvement in these trends, from making spectacles of hosting tech executives to driving and publicly endorsing major buildout projects. Polls don’t yet include ‘who do you trust to defend American interests against AI developers’, but once they do, I struggle to see how the GOP would perform well on it given recent headlines and enduring positioning. I think GOP strategists might see that trend and shape their strategy accordingly.
Cross-tab make jobs risk worse for Democrats
We don’t know all that much about the exact shape of AI labor market impacts just yet. But directionally, some trends seem clear: white-collar work is more exposed than blue-collar work, indicating greater exposure of urban populations. Junior roles are more exposed than senior roles, indicating greater exposure of younger populations. If you map these trends to electoral data, you get a fairly polarised picture: Democratic electorates are much more exposed. I’m unsure how much this matters: the key political driver is usually not that much job loss, but anxiety about job loss – and it’s unclear how closely that will track the details of that effect. But a low resolution representation of these trends (‘AI takes away bullshit jobs’, for instance) could stick across the electorate.
Once that happens, the incentive for Republicans to cater to job worries goes down, the incentive for Democrats goes up. This creates asymmetric electoral pressure: Democrats will face anxious constituencies demanding action, while Republicans can more easily dismiss AI displacement as affecting ‘other people's jobs’ in blue cities. The GOP base may even see some schadenfreude in coastal professional displacement, reducing pressure on Republican politicians to offer protective policies. Meanwhile, Democrats risk looking out of touch if they don't respond to their voters' economic anxieties with concrete anti-AI measures - of which there are notably few that actually work, making matters even worse.
There might be a winning pro-AI GOP narrative
The recently released AI Action Plan contained the makings of a direly needed alternative GOP message on AI and jobs: Present labor effects as a trade-off, in which the US gains blue-collar employment in AI infrastructure but loses white-collar work. I’m unsure whether this is effective policy, but what matters here is the narrative. And a narrative can be told even through a small nominal impact, with headline news about some jobs created through datacenters, images and campaign events and speeches hammering the message home. This is politically attractive for three main reasons:
First, it helps with the incumbency problem: It provides headlines and examples of a boost in employment just as the 'creative destruction' wave of automation hits. It allows shifting the story from 'AI is bleeding out the economy' to 'AI is changing the economy', which is much less of a political risk.
Second, it reinforces the cross-tabs point from above. Shifting jobs from young urban white-collar workers to blue-collar workers in red states is a net positive for the GOP - your local losses among the former probably don't hurt too much, and gains among the latter reflect your agenda. That could even make some job disruption a positive for the GOP: 'Displacing coastal elites in bullshit jobs' as a message fits a lot of the administration’s current rhetoric and gives the base a way to remain unconcerned about AI impacts, even in the face of general displacement stories.
Third, it ties in favourably with existing administration commitments around trade and economic policy, which already introduces the point that manufacturing jobs are particularly valuable to the US. The job shift story cashes in on the same framing required to sell tariff policy.
If this story turns out to resonate well – first with GOP policymakers, then with the electorate –, it can provide a ‘way out’ of the GOP’s conundra around factional conflicts and job disruption exposure. Any such policy that can keep the volatile coalition together and reduce exposure to a powerful Democratic message seems likely to be picked up swiftly and broadly.
Outlook
What does that mean for AI policy? First, I think it makes the entire story of labor-impact-related policy a bit more complicated. If your policy planning is banking on an emergence of political support for strong AI regulation based on labor market impacts, I think that would be mistaken. Just because jobs are an important political issue, it’s not obvious that parties will agree that AI regulation is the way to win on AI and jobs. Quite the opposite: Especially for the GOP, a lot hinges on ‘AI jobs mean regulation is required’ not becoming the sentiment of the day.
Second, I think predictions open some niches for successful and effective messages. For instance, if you recognise that the GOP is in need of a strong absorption story, that’s a great touchpoint for scaling up effective reallocation mechanisms, intermediary insurances, and strategic support of absorptive sectors. And if you recognise the particular overexposure of the Democrats to AI job disruption and therefore their perhaps likely tendency to pivot against AI, you might want to get into that conversation early: through conversations with involved labor stakeholders, and with representatives from particularly exposed districts. Whether you want to use that to preempt explosive backlash through less invasive policy, as a launchpad for safety policy, or to actually address the displacement problem, I’ll leave up to you.
I’ll close on this: I’m glad public debate has correctly identified the political potential of AI job disruptions. I think to make this insight useful for policy planning and strategy, we should do a better job examining how vague political sentiment translates into party strategy and policy preferences. The story of populism entering the AI debate will be a bit more complicated than frequently given credit for.