Anthropic launches Public Record survey: 64% of Americans fear AI job loss, only 15% trust AI companies

Anthropic on June 12 released the inaugural results from its Anthropic Public Record, a new survey series the company says it will repeat to track how U.S. public attitudes toward AI shift over time. The first wave surveyed nearly 52,000 Americans across all 50 states between November and December 2025.

What's new

The survey captures a public simultaneously drawn to AI's potential and worried about its consequences.

Top hopes:

• 48% of Americans ranked curing diseases like cancer or Alzheimer's among their top three hopes for AI
• Assisting people with disabilities followed at 36%

Top fears:

• AI-induced job loss was the most common fear in every state, held by 64% of Americans
• Cognitive dependency: 56%
• Misinformation: 52%

Trust and governance:

• Only 15% of Americans trust AI companies to make decisions about how AI is developed and used — lower than any other tested institution, including government and international bodies
• Over 70% of Americans surveyed believe the government should play a role in regulating AI, on a bipartisan basis
• Top priorities for government action: privacy protection (56%), child safety (52%), and corporate liability (49%)

Anthropologic also identified a distinct "Integrated User" segment — the 6% of Americans using AI daily for both work and personal tasks. This group shows lower concern about harms but maintains similar support for government oversight as the general population, suggesting familiarity reduces fear without reducing the appetite for governance.

Context

The Public Record initiative arrives at a crowded moment for Anthropic. In the past two weeks, the company released Claude Fable 5, submitted a confidential S-1 to the SEC, launched Claude Corps (a $150 million nonprofit fellowship program), and proposed a new AI governance framework tied to a 10²⁵ FLOP compute threshold. Publishing a large-scale public attitudes survey layers a transparency and accountability posture on top of those moves.

The 15% trust figure stands out. It arrives weeks before Anthropic's IPO preparation becomes more public-facing — and gives regulators in the U.S. and EU data to cite when arguing that the public demands independent oversight rather than industry self-governance.

Anthropologic said it plans to publish results from each subsequent wave, allowing year-over-year comparisons as Claude Fable 5 and future models become more embedded in everyday work.

Why it matters

The Anthropic Public Record is unusual among AI company communications. It is not a benchmark, a product launch, or a safety card — it is a direct attempt to surface what ordinary people, not developers or enterprise customers, actually want from AI and who they trust to guide it.

Three signals are worth tracking in future waves:

1. The trust deficit is structural. At 15%, AI companies rank below every other tested institution. That is a ceiling for voluntary adoption in regulated sectors — health care, government, finance — where procurement decisions hinge on public legitimacy as much as technical performance.

2. Job displacement remains the dominant fear. Despite years of messaging about AI augmenting rather than replacing workers, 64% of Americans in every state flag job loss as their primary concern. That gap between industry framing and public perception is not shrinking on its own.

3. Integrated Users diverge on fear, not on governance. The segment that uses AI most intensively wants guardrails at nearly the same rate as the general public. If that pattern holds as adoption grows, it suggests demand for AI oversight will not erode as familiarity increases — a data point that complicates arguments that more usage will naturally produce more trust.

Anthropolic has committed to making the Public Record a standing instrument. The first comparative wave will likely be the more consequential dataset.