On paper, Leading the Future walked away from the 2026 midterms looking like a winner. The sprawling pro-AI super PAC network, financed by some of the wealthiest names in Silicon Valley, still had more than $20 million sitting in the bank. Dozens of congressional candidates it backed had won their races, and those victories spanned both parties, a feat that would make any political operative proud. At first glance, that is exactly what you want from a machine built to shape Washington. But political victories are rarely measured by bank balances alone, and the details underneath the headline numbers tell a far more complicated story. Most of the money went to candidates who were never in serious danger of losing, safe incumbents and anointed successors in districts where the partisan outcome was all but predetermined. In the handful of genuinely risky races where Leading the Future actually chose to test its strength, the results ranged from underwhelming to outright embarrassing. The network did not get the feared candidates elected, and it did not manage to bury the ones it opposed. In the end, it spent a fortune to purchase influence in places where influence was already cheap, while failing to prove it could move the needle where the outcome actually mattered. It is the political equivalent of buying a lottery ticket and leaving the winning numbers in your pocket.
To understand why that is such a striking outcome, you have to remember how terrifying Leading the Future looked when it first arrived on the scene. The network launched in August 2025 with a war chest of around $100 million, backed by some of the most powerful names in technology: the venture capital giant Andreessen Horowitz, Palantir cofounder Joe Lonsdale, and OpenAI’s Greg Brockman, among others. It was not a ragtag startup; it was staffed and advised by major political strategists with deep experience in national campaigns. The group described itself as a “structured ecosystem” designed to give donors “maximum flexibility to engage in the 2026 election cycle and beyond,” and the structure was genuinely impressive. It included 501(c)(4) organizations, which do not have to disclose their donors, as well as federal and state super PACs, and it deliberately catered to both sides of the aisle. American Mission was built largely to support Republicans, while Think Big was built largely to support Democrats. The whole apparatus was designed to make it impossible for any single party or committee to capture the network, and the overlapping strategists and shared donor lists immediately drew comparisons to Fairshake, the massive pro-crypto super PAC network that dominated the 2024 election cycle with hundreds of millions of dollars in spending. For people watching from Washington, the message was unmistakable: Silicon Valley had learned the crypto playbook, and it was coming for the AI policy debate with the same ruthlessness, the same money, and the same willingness to play hardball in primaries as well as general elections.
The fear, at least among AI safety advocates and tech accountability groups, was visceral and justified. Leading the Future’s public promise to “oppose policies that stifle innovation” sounded reasonable enough on its face, but the way the network went about it suggested something more menacing. It did not simply want to fund allies; it wanted to intimidate enemies. Sacha Haworth, executive director of the Tech Oversight Project, described the effect in stark terms. “They came out guns a-blazing, saying, ‘Be afraid, we are going to spend our millions to influence not only how lawmakers legislate us but also against you in your primaries and in your reelection,’” she told WIRED. “The purpose of that was to bully lawmakers into acquiescence, intimidate them into not sticking their necks out on AI safety or tech industry regulation at all.” In other words, the network was not just trying to elect people who shared its vision. It was trying to create a chilling effect so powerful that even moderate Democrats and Republicans would think twice before supporting any kind of AI oversight. A legislator might genuinely believe that powerful models need guardrails, but if opposing Leading the Future meant facing a multimillion-dollar onslaught of attack ads in the next primary, the calculus shifted dramatically. That is the kind of fear that does not show up in campaign finance reports, but it can be the most powerful force in politics. And for a while, it seemed to be working. Lawmakers on both sides of the aisle had to know that the AI industry was watching, and that the industry had both the resources and the will to punish anyone who crossed it.
The clearest test of that strategy came in New York’s 12th congressional district, where Leading the Future took on Democrat Alex Bores. Bores was not some unknown outsider; he was a former Palantir engineer and a sitting state lawmaker who had made a name for himself as a thoughtful voice on technology policy. But he was running on an AI safety platform, which made him an existential threat to the network’s vision of a deregulated, innovation-first future. Leading the Future poured millions of dollars into attack ads against him, hoping to crush his primary bid before it could gain momentum. The expectation, according to Haworth, was total annihilation. “They fully expected to bury him, put his head on a spike, and parade it around as proof that they could,” she said. But the strategy backfired spectacularly. Instead of fading into obscurity, Bores became the center of a high-profile proxy war between competing visions of artificial intelligence. The pro-AI safety network Public First, which has significant financial backing from Anthropic, stepped in and spent $15 million boosting his candidacy. Suddenly, the race was no longer a local primary; it was a national referendum on the future of AI policy, with two Silicon Valley factions spending massive sums to prove which philosophy would dominate Washington. Bores did not win the nomination, but he came in second in a crowded field, and the victor, Micah Lasher, had also campaigned on tech accountability and AI safety. In a single race, Leading the Future had demonstrated both its power and its limits. It could spend enough money to make the race ugly, but it could not force the voters to embrace its agenda. It could not even prevent the two strongest candidates in the district from running on the very message it was trying to suppress.
When you look at the broader ledger, the picture becomes even more damning. Leading the Future’s donations went overwhelmingly to safe candidates, people who were going to win their elections with or without the network’s help. There is a certain logic to that kind of spending: it buys loyalty, builds relationships, and ensures that friendly lawmakers feel obligated to take your calls when they get to Washington. But it does not prove that the network can actually shape an election. The handful of races where Leading the Future took a real risk and went up against a serious opponent or a contested primary did not go according to plan. The network’s supposed invincibility was built on the idea that it could make or break careers, that no lawmaker or challenger could survive a multimillion-dollar assault from the AI industry. That reputation was always more important than any single race, because it was what made politicians nervous enough to preemptively soften their positions on regulation. But after the 2026 midterms, that reputation is in tatters. A super PAC that spends its money on safe incumbents is not a tidal wave; it is a sprinkler. It waters the lawn in places that were already green while leaving the actual battlefield dry. The results suggest that Leading the Future’s real strategy was less about winning tough fights and more about projecting strength, and when it finally stepped into the arena with something on the line, it could not deliver. That is not the kind of failure that shows up on a balance sheet, but it is the kind that matters most in politics. Candidates and donors notice when a machine promises to bury its enemies and ends up with a second-place finish instead.
So what does this all mean for the future of AI politics? It would be a mistake to declare Leading the Future a total failure. The network still has significant money in reserve, it has a roster of successful candidates who may feel grateful for its support, and it will almost certainly be back for the next election cycle with a new strategy and a renewed sense of purpose. But the midterms exposed a fundamental truth about money in politics: it is a megaphone, not a mandate. You can use it to amplify a message, but you cannot use it to make people believe something they do not already believe. The Bores race proved that AI safety has a real constituency, one that is willing and able to fight back with its own money and its own messengers. It also proved that voters are not simply puppets who dance when a super PAC pulls the strings. They have their own concerns about technology, their own fears about concentrated power, and their own willingness to support candidates who take those fears seriously. Leading the Future may have wanted to make the 2026 elections a demonstration of Silicon Valley’s dominance over AI policy, but what it actually demonstrated was the resilience of democratic choice. The network’s aura of invincibility is gone, and every lawmaker who watched it fail to bury Alex Bores knows that the empire has cracks. That does not mean the next cycle will be easy for AI safety advocates, or that Leading the Future will not come back with a sharper, more disciplined approach. But it means that the debate is alive, that the money is not the only thing that matters, and that the future of AI policy will be fought in the open, between real candidates and real voters, not just in the shadows of dark-money committees. In the end, the biggest winner of the 2026 midterms may not have been any candidate or super PAC, but the simple, stubborn idea that elections still belong to the people.
