Nine AI leaders, nine majorities of distrust: the floor onboarding copy cannot lift

CNBC and the Generation Lab, a Washington research company co-founded by Cyrus Beschloss, asked 18-34-year-olds whether they trust each of nine named figures in the AI industry to act responsibly on AI, and in every single case a majority said they do not ^[1][2][3]. For anyone shipping an AI feature to consumers, that is the ambient condition of the market: the trust question is answered before the first screen renders.

The striking part is not the height of the numbers but how little they vary. Microsoft's Satya Nadella performed best of the nine, at about 65% distrust, which still means roughly two in three respondents ^[4][23]. Mark Zuckerberg, Jensen Huang and Elon Musk sat at about 70% each ^[5][6][7]. Sundar Pichai, Dario Amodei and Sam Altman came in at about 75% ^[8][9][10]. Palantir's Peter Thiel drew about 79% and Alex Karp about 81% ^[11][12]. The entire industry's public leadership fits inside a 16-point band ^[22].

Two details in that spread should interest product teams more than the headline. First, Amodei and Altman land on the same reported figure ^[21]. Whatever differentiation exists between how those two companies talk publicly about responsibility, it is not producing a distinguishable trust number in this cohort. Second, the ordering does not track how much consumer AI a company actually ships, which suggests the survey is measuring a general disposition toward tech leadership rather than a judgment on any specific product decision.

The survey also captured what young respondents expect AI to do to them. Forty-five percent said its impact on their careers will be negative and 10% said positive, a ratio of four and a half to one against ^[13][19]. That is the prior your copy is arguing with. A tooltip promising to save the user time is being read by someone who has already concluded the technology is coming for their earnings, and no amount of friendly onboarding language reframes that.

Context matters for how much weight to put on the AI findings. Nearly 80% of the same respondents expressed negative sentiment about the US economy, 50% expect it to get worse against 22% who expect improvement, and 46% hold a favorable or very favorable view of democratic socialism against 23% who view it unfavorably ^{[14][15][16][20]}. The AI distrust sits inside a broader pessimism, which cuts both ways for operators. No single trust incident created these numbers, and no single trust-building campaign will move them.

The practical consequence is that consent design stops being a compliance exercise and becomes the product. Default-on data sharing, retention windows, model-training toggles, and where the AI output sits relative to the user's own work all get read by an audience predisposed to assume bad faith. Teams that assume neutral-to-curious users when setting defaults are mispricing the risk of a backlash.

Handle the figures with care. As reported by Gizmodo, the distrust numbers are given as approximations rather than exact percentages, and the write-up does not carry a sample size, field dates, question wording, or a margin of error ^[17]. Beschloss has appeared repeatedly on CNBC in this explanatory role, including a year ago, so comparable waves may exist for anyone tracking the trend ^[18].

Two things worth watching: whether a future wave shows any leader breaking below majority distrust, and whether stated distrust shows up in behavior. This survey measures what young people say about AI leaders, not what they opt out of.