The AI Paradox: Usage is Going Up as Trust is Going Down

21% of American adults say they trust AI most or all of the time, while 76% say they trust AI either hardly ever (27%) or only some of the time (49%). AI usage has risen — 51% have used AI to research topics (up from 37% in April 2025), 27% use it for school/work (up from 24%), and 24% have used it to create images (up from 16%); the share who never used AI fell to 27% from 33%. Concern is growing: 55% now say AI will do more harm than good in daily life (versus 44% in 2025), 70% expect AI to decrease employment opportunities (56% in 2025), only 7% foresee job increases, and 81% of Gen Z predict decreased opportunities.

Analysis

The widening gap between rising use and falling trust in AI is a structural headache for ad-funded consumer platforms: advertisers will pay more to avoid brand risk, and that reallocates budgets toward environments with stronger content provenance and measurement. Expect programmatic CPMs to bifurcate — premium, verifiable inventory (first‑party, walled gardens, verified publishers) will hold or rise, while open-web generative content inventory suffers persistent discounting unless verification tools scale quickly.

Regulatory and political cycles amplify this: in an election year regulators and lawmakers have incentives to spotlight perceived harms, which raises compliance costs (audit trails, disclosure features, human review) that hit margins of smaller, ad‑dependent players faster than large vertically integrated platforms. The likely second‑order winners are vendors selling auditability, identity resolution, and content attribution — these are de‑risking purchases advertisers will accept during budget re‑allocations.

For Snap specifically, the mechanics are clear: margin risk from higher content moderation and ad efficacy churn, offset by potential upside if its AR/AI features measurably lift engagement and first‑party signal capture. The tradeoff is binary within ~6–12 months — either advertisers demand premium attribution and punish open inventory, or product-led monetization improvements justify multiple expansion.

A contrarian cushion: user habituation to AI still increases raw engagement and data flow, so platforms that can rapidly convert that signal into verifiable first‑party targeting gain disproportionate pricing power. The market may be overpricing structural ad flight relative to the timeline required for regulatory changes and advertiser contract shifts (12–24 months).

Key Decisions for Investors

• Short SNAP via a 6‑9 month put spread sized to 1–2% portfolio risk: sell a put 10% OTM and buy a put 20% OTM (ratio depending on skew) — thesis: midterm regulatory headlines + advertiser budget reallocation compresses multiples by 10–20% within 3–9 months; target 2:1 downside capture vs premium paid.
• Pair trade (6–12 month horizon): short SNAP equal‑notional / long GOOGL (or META) — capture rotation from open-web, generative‑risk inventory to walled gardens with stronger measurement; expect asymmetric payoff if advertisers shift conservative spends, target 15–25% relative outperformance.
• Long cybersecurity/verification vendors (examples: CRWD, PANW) via 12‑month calls or outright small longs — buys exposure to increased spending on provenance and auditability; aim for 20–40% upside if regulatory guidance mandates provenance standards within 12 months.
• Tactical contrarian long: small, time‑limited SNAP call position (3–6 months) only after evidence of product monetization (e.g., new first‑party ad product rollouts or improved AR metrics) — risk limited exposure to capture a fast, product‑driven re‑rating event; use sell‑to‑close triggers at +40–60%.