A strong majority of AI-engaged professionals believe Anthropic is right to hold the line on autonomous weapons and mass surveillance, even at the cost of government contracts.
But when it comes to AI-driven layoffs, the same audience is split between urgency and skepticism, and most say they haven't felt the impact on their own teams yet.
This week's AI Pulse survey drew 91 responses from listeners of The Artificial Intelligence Show. The results capture a striking tension: people broadly support AI safety red lines, broadly believe AI layoffs are real, and broadly report that none of it has touched their own teams. The gap between belief and experience is the story.
As always, this is an informal poll of our podcast listeners, not a scientific study. The respondents skew AI-aware and professionally engaged with the technology. That context matters when interpreting the numbers.
We asked: Where do you stand on the Anthropic-Pentagon dispute over AI safety red lines?
The results were decisive:
61.5% said Anthropic is right to hold the line on autonomous weapons and mass surveillance, even at the cost of government contracts.
When you combine the first two responses, nearly 79.1% of respondents agree with Anthropic's actual red lines, meaning the disagreement is only about tactics, not principles. Just 4.4% believe AI companies should give the government unrestricted access, which is the position the Pentagon effectively took when it designated Anthropic a supply chain risk.
The 16.5% who see this as a political power play is worth noting. As the situation has escalated, with a leaked internal memo from Dario Amodei, Sam Altman apologizing to OpenAI staff, and reports that Claude is still powering military operations in Iran, the political dimensions of this dispute have become harder to separate from the safety ones.
We asked: Block cut nearly half its workforce this past week and named AI as the reason. What's your reaction?
44.0% said the layoffs are real, but the pace will be slower than the headlines suggest.
29.7% said it's mostly a correction from pandemic over-hiring — AI is a convenient narrative.
25.3% said this is the beginning of a major wave of AI-driven layoffs across industries.
1.1% said they're already seeing AI reduce headcount or roles at their own organization.
Combine the 44.0% who believe layoffs are real (but slower) with the 25.3% who see a major wave beginning, and 69.3% of respondents believe AI-driven job losses are genuinely happening — not just a convenient excuse for pandemic-era corrections.
That said, the largest single group (44.0%) tempers that belief with realism about pace. This aligns with what SmarterX founder and CEO Paul Roetzer discussed on Episode 201: the technology capability is advancing rapidly, but organizational adoption, and therefore the actual displacement, moves on a much slower timeline.
The near-absence of respondents who are personally witnessing AI-driven cuts at their own organizations (just 1.1%) sets up a tension that the third question makes explicit.
We asked: How has AI affected the size of your team or department in the last 12 months
51.6% said no noticeable impact yet.
34.1% said headcount has stayed the same, but roles are shifting.
13.2% said headcount has shrunk — some positions weren't backfilled or were eliminated.
1.1% said headcount has grown — they're hiring more people with AI skills.
This is the most revealing question of the three. Nearly 70% of respondents believe AI layoffs are real and happening (Q2), yet 51.6% report no noticeable impact on their own team. That's the gap between the macro narrative and the micro reality — and it's exactly the dynamic that makes this moment so disorienting for professionals trying to gauge how urgently they need to act.
The 34.1% who report roles shifting without headcount changes may be the most important signal here. This is the quiet middle stage of AI adoption: no one gets fired, but the nature of the work changes. Job descriptions evolve. New skills become expected. The transformation happens beneath the surface before it becomes visible in headcount numbers.
Meanwhile, 13.2% are already seeing headcount shrink, with positions not backfilled, roles eliminated. That's roughly 1 in 8 respondents. Not a wave yet, but not nothing.
In our ongoing AI Pulse surveys, we gather insights from listeners of our podcast to get a sense of how our audience feels about various topics in artificial intelligence. Each survey is conducted over a one-week period, coinciding with the first seven days after an episode is released. During that time, our episodes typically receive around 11,000 downloads.
Our survey results reflect a self-selected sample of listeners who choose to participate, and typically we receive a few hundred responses. While this is not a formal or randomized survey, it offers a meaningful snapshot of how our engaged audience perceives AI-related issues.
In summary, when you see percentages in our headlines, they represent the views of those listeners who chose to share their opinions with us. This approach helps us understand the pulse of our community, even if it doesn't represent a statistically randomized sample of the broader population.