AI Is Selling a Story. The Budget Math Doesn't Back It Up, and Consumers Don't Trust It Either.

Since earlier this year, AI has become Washington's favorite fiscal fantasy and Madison Avenue's newest buzzword. Both relationships are showing cracks.

The Budget Pitch Doesn't Pencil Out

The argument from tech evangelists, including Elon Musk, is that AI-driven productivity growth will generate enough tax revenue to rescue the federal budget. Musk has specifically called AI and robotics "the only thing that can solve for the debt situation."

According to analysis published by Reason, even a genuinely impressive AI-driven productivity boom, comparable to the late-1990s tech surge that lifted annual productivity growth about 1 percentage point above trend, would raise federal tax revenues by $143 billion annually by 2028, climbing to $834 billion annually by 2036.

That sounds large. Against a projected $4.4 trillion annual deficit under current policies, it covers roughly one-fifth of the gap. One-fifth.

Reason's analysis goes further: even if AI delivered a 2 to 3 percentage point sustained boost to productivity growth, an outcome beyond what most credible economists forecast, it still would not come close to balancing the federal books. The underlying fiscal problem is structural: Social Security and Medicare spending trajectories that no technology wave has ever been able to legislate around.

That productivity gain would be partly offset by headwinds already baked in. Reason notes that labor force growth is already slowing due to falling fertility rates, retiring baby boomers, and immigration restrictions. This means AI gains would partly replace lost growth rather than add on top of a healthy baseline.

The honest version of the AI-fiscal case is this: AI could generate real revenue and real savings at the margin. It is NOT a substitute for the hard choices on entitlement reform and taxation that politicians of both parties have refused to make for decades. Betting the U.S. fiscal future on a technology breakthrough is the same category of wishful thinking as betting it on taxing billionaires or cutting foreign aid. Directionally real, numerically insufficient.

Consumers Are Tuning Out the AI Label

While Washington is being sold on AI's transformative potential, consumers are pushing back on AI in a different arena: brand messaging.

A survey of 2,000 respondents released June 16, 2026 by WordPress VIP, the enterprise arm of Automattic which owns WordPress, found that 60% of U.S. consumers say brands that use AI in their messaging are a turnoff. The survey covered 1,200 U.S. adults and 800 enterprise decision-makers and CMOs, conducted in April 2026, according to TechCrunch.

86% of respondents said they don't fully trust AI-generated answers and still want to check original sources. 42% said AI-generated answers without clear attribution are trusted less than airline fees, confusing privacy policies, and medical bills. Those are categories specifically designed to frustrate consumers.

Nearly three in four respondents told WordPress VIP the internet feels less human than it did a decade ago.

The strongest counterpoint to this consumer wariness: it may not actually change behavior at scale. The same report found that 60% of enterprise respondents reported AI search engines and answer platforms drove more traffic to their sites over the past year. Consumers say they distrust AI; they are also clicking on AI-surfaced results. The gap between stated preference and actual behavior is a real complication for anyone drawing firm conclusions from survey data alone.

WordPress VIP's CTO Brian Alvey framed the dilemma plainly: "People used to build websites for other people. Now you have to build websites for AI agents acting on behalf of those people." He added that content not legible to AI systems is effectively invisible to a growing share of how people search.

74% of enterprise decision-makers in the survey said AI discoverability and attribution are a main or significant priority, even as their customers express distrust. Brands are caught between two audiences: the AI systems that surface their content, and the humans who are growing more skeptical of that surfacing.

What States Are Actually Doing About AI

While the federal government debates AI's fiscal magic and brands chase AI visibility, state governments have been quietly legislating. According to the National Conference of State Legislatures, all 50 states, Puerto Rico, the Virgin Islands, and Washington D.C. introduced AI-related legislation in 2025. Thirty-eight states enacted or adopted roughly 100 measures.

The range is wide. Arkansas clarified AI-generated content ownership. Montana passed a "Right to Compute" law placing requirements on critical infrastructure controlled by AI systems, while also prohibiting government restriction of private computational resources except under compelling government interest. New York now requires state agencies to publicly inventory their automated decision-making tools. New Jersey adopted a resolution urging AI companies to voluntarily protect employee whistleblowers.

None of that legislation touches the fiscal question Reason raises. It is process and accountability work, real and necessary, but operating at a different level than the macro-budget claims Musk and others are making in Washington.

The Unresolved Question

The WordPress VIP report notes that 80% of consumers believe web information should remain openly accessible rather than controlled by a small number of large organizations. That preference sits in direct tension with the current trajectory: a handful of AI platforms becoming the primary filter through which people access information, with the web's original-source ecosystem shrinking in visibility.

Whether that tension produces meaningful regulatory action or whether it stays a survey preference that consumers quietly ignore remains an open question as of June 16, 2026. The NCSL database shows states moving faster than Congress on AI accountability rules, but none of the state measures enacted so far address the concentration-of-access problem the WordPress VIP survey identified.