AI will be the defining issue of the 2028 election, and the sentiment data is already pointing that way

Dario Amodei, chief executive of Anthropic, states flatly that AI will be “maybe the most important issue in the presidential election in 2028.” That is not a hedge or a conversational aside. It is a timeline, and the public sentiment data is beginning to confirm the direction if not yet the magnitude.

Nathan Labenz, who tracks AI adoption closely, reports that for the first time a majority of adults under 30 say they are more concerned than excited about AI. Their concern now matches that of people in their 30s, 40s, and those 65 and older. The only cohort still tilting toward excitement over concern is the 50-to-64 age group. That shift among younger adults is significant precisely because they were, until recently, the demographic most reliably enthusiastic about new technology. The sentiment movement is consistent with the political trajectory Amodei is describing.

The concern is not uniformly distributed across countries, which itself carries political weight. David Sacks, who served as the US AI and crypto czar, puts the contrast plainly: AI optimism in China sits above 80 percent on the question of whether AI will be more beneficial than harmful, while the equivalent figure in the United States sits around 30 percent. That gap, a divergence of roughly 50 percentage points, is not merely a cultural curiosity. It signals that the two countries are approaching the same technology from entirely different political starting points, with consequences for regulation, investment, and public tolerance of disruption.

For the first time, a majority of adults under 30 say they're more concerned than excited about AI. Their concern is now on par with those in their 30s and 40s and those 65 and up. And so we have the only group still under is the 50 to 64 group. So the Gen Xers are still majority are still more excited than concerned. Nathan Labenz

The timelines attached to that disruption are specific enough to demand attention. Ryan Greenblatt, an AI safety researcher, places full automation of AI research and development at around 2030 or 2031, with AI beating all humans across jobs reaching a median expectation of around 2033. Amodei separately projects a cash dividend to all Americans in 2033. Bernt Bornich states that hard takeoff, meaning robots building robots, chip fabs, data centers, and handling mining and refining autonomously, is less than a decade away. These are not identical predictions, but they cluster around the same window, and that window runs directly through the next two presidential cycles.

The anxiety is not confined to expert discourse. Melisa Tokmak, who works in hiring, reports that Gen Z candidates are arriving with what she describes as a “permanent underclass mentality,” obsessed with the idea that if they do not make their money in the next 18 months, AI will subsume their economic value. That framing, whether or not it is calibrated correctly, is shaping real career decisions right now. When a technology begins reorganizing how young workers think about their futures, it has crossed into mainstream political territory regardless of what policymakers choose to do with it.

The geopolitical dimension is accelerating that politicization. Harry Stebbings reports that Chinese open-source models have improved rapidly, and that a model he refers to as “Kimmy K3” recently beat the best closed-source American models on what he calls a pretty important subset of tasks. Jason Calacanis, a technology investor, puts the US lead over China in blunter terms: possibly months, if that. Eric Vishria adds that China is bringing on ten times as much energy next year as the United States, a figure that matters because energy is the proximate constraint on AI compute. These pressures are already entering the language of industrial policy, and they will be fully legible to voters by 2028.

What makes Amodei’s prediction about 2028 credible is not that he is a prominent figure in the industry. It is that the forces he is pointing to, shifting public sentiment, widening international divergence, concrete near-term economic disruption, and a technological timeline now visible to ordinary workers, are all already operating. The question for 2028 is not whether AI will be on the ballot in some abstract sense. It is which candidate or coalition will offer the more coherent account of what to do about it, and whether the institutions currently responsible for that question are moving fast enough to have an answer ready.