President Donald Trump has placed a major bet on expanding artificial intelligence across the United States, pushing hard for more investment, bigger infrastructure projects, and fewer rules. But some experts now believe this strategy could seriously backfire, especially if what they see as an AI investment bubble finally bursts and the public grows more hostile toward the technology as its risks become harder to ignore.
That warning comes from AI researcher Gary Marcus, whose prediction appeared in a feature by Politico Magazine that explored 15 possible “black swan” events unexpected developments that could dramatically shape the future and shock the world.
Marcus, a retired professor from New York University and a longtime AI entrepreneur, suggested that by the end of 2026, Trump may start quietly pulling away from the aggressively pro-AI stance that defined his approach in 2025.
According to Marcus, the massive AI infrastructure projects Trump promoted after taking office could end up looking like costly mistakes expensive, underused, and unable to generate the promised returns.
He also argued that Trump’s failure to properly regulate AI, despite concerns shared by voters and politicians on both the left and the right, would become increasingly hard to defend.
Marcus warned that as problems linked to AI pile up, Trump would face blame from all sides. These issues could range from deepfake videos and misinformation, to people experiencing mental health crises triggered by chatbots, to large-scale cyberattacks powered by AI tools.
On top of that, public anger is expected to grow over the spread of massive data centers, rising energy costs, and what many see as greedy behavior by powerful AI companies.
He added that AI stocks could crash, and the technology that Silicon Valley once celebrated as its next big breakthrough may start to look more like a fadan expensive solution searching for a real problem, with business models that simply don’t make financial sense.
If that happens, Marcus believes Trump would quickly distance himself from the entire project, abandoning it as soon as it becomes politically inconvenient.
According to Marcus, Trump would not hesitate to walk away once the mood shifts, acting as though the whole AI push was never central to his agenda in the first place.
There are already signs, he argues, that this turning point may be approaching. Public opinion polls show growing skepticism about expanding AI, and protests have begun to break out across the country over the construction of new AI data centers.
At the same time, Trump’s efforts to weaken state-level regulation of AI have sparked resistance even within his own party. Figures such as Ron DeSantis have pushed back, signaling that concerns about AI are no longer limited to one political side.
Taken together, these developments suggest that the AI boom Trump embraced so enthusiastically could soon become a political liability rather than a strength, with consequences that follow him well beyond the next election cycle.
