Last week, Oren Cass told readers of The New York Times that Big Tech’s oldest trick, build first and ask permission never, has stopped working on AI the way it worked on smartphones and social media. I’ve personally known Cass a long time, so when he writes about a technology industry that can no longer talk its way out of a trust problem, I don’t just read it as commentary. I check it against what’s happening in our own town, because ABRSD finished its own AI accountability document several months before Cass made his argument in print.
The district’s guidebook won’t make headlines outside of our local community. It should. It’s a working answer to the exact question Cass says the industry keeps dodging, and buried inside a PK-12 policy document is a governance template that translates directly into what earns brand citations in AI search results right now.
Adoption Is Holding, Trust Is Not #
Cass’s argument is that the political formula Silicon Valley ran for two decades, promise the benefits, ask regulators to wait, move on to the next product, has stopped converting. His evidence is specific and current. An August 2026 Annenberg Public Policy Center survey found 61% of Americans opposed new data centers in their own area, up from 49% just months before, with opposition crossing party lines. Annenberg’s data also shows 68% think government regulation of AI has been too weak, not too aggressive.
None of that reads as anti-technology. Cass is careful about that distinction, and it’s the right one. People use the chatbots daily while opposing the data centers built to run them. What’s collapsed isn’t demand for the product. It’s belief that the companies selling it will do right by the people affected by it.
That gap between usage and trust is the same gap I’ve been tracking all year in YouGov’s brand research and interviews with Reuben Staines. AI brands are winning consideration. They are not winning trust. Cass just gave that finding a much bigger, more political stage than a marketing column usually gets, and the stakes he’s describing, tax subsidies, grid capacity, entry-level jobs, are a lot higher than click-through rate. But the underlying mechanism is identical. Adoption without accountability produces exactly this kind of backlash, whether the audience is a swing voter or an in-market buyer.
1 District, 5 Principles #
Here’s what Cass’s column doesn’t have, and what ABRSD’s “AI Guidelines & Guardrails” does. A concrete, published example of an institution building the accountability structure before the backlash instead of after it.
ABRSD’s guidelines rest on five principles:
- Humans First.
- Adaptive Literacy.
- Responsible Stewardship.
- Rigorous Governance.
- Intentional Use.
Skip past the education-specific language and what’s left is close to a corporate AI governance charter. Under Rigorous Governance, vendor contracts must guarantee student and staff data is never used to train commercial large language models. Under Intentional Use, every piece of AI-generated content, instructional material, or external communication has to clear human review before it goes out, a requirement the guidebook calls the Human-in-the-Loop rule. Under Responsible Stewardship, staff are expected to disclose their own AI use and teach the technology’s environmental and intellectual-property costs alongside its benefits.
A district survey of its own high schoolers, taken in March 2026 and included in the guidebook, backs up that the framework is landing. 79% of students said they understand when using a GenAI tool would let them skip work they actually need to do to learn. 72% said their teachers are clear about when AI use is and isn’t allowed. Those numbers exist because the district wrote the rule down, named who’s accountable for it, and asked students afterward whether it held up. That’s the evidence-over-promises model Cass says Big Tech has never had to build, produced by a 16-person working group of teachers, students, administrators, and community advisors, on a budget nobody would call generous.
A public school district with none of Big Tech’s resources produced a more credible accountability document, in a matter of months, than most AI companies have managed in years of promising it’s coming.
How To Put This To Work #
If you run content, brand, or search strategy and you’ve been treating “we use AI responsibly” as a marketing line rather than a document, ABRSD’s structure gives you three things to lift directly. First, publish a named, dated AI-use policy, not a vague ethics statement, and make it indexable. Treat it as content in its own right. A citable primary-source policy page is exactly the kind of asset that AI Overviews and answer engines pull from when a user asks whether a brand handles data or AI responsibly, and right now almost nobody in your category has written one.
Second, put a Human-in-the-Loop line on every AI-assisted asset you publish and mean it. Disclosure of AI involvement, backed by an actual named reviewer, is a potential trust signal with both readers and the LLMs now citing your content, not a legal hedge to bury in a footer.
Third, state plainly, the way ABRSD did for vendor contracts, whether customer or user data feeds any third-party model training. Cass’s numbers show people are done taking that on faith from any company. Answering the question before it’s asked is worth more, this year, than answering it well after you’ve been caught not answering it.
Big Tech is going to keep losing this argument in Washington because it keeps trying to win it with a pitch instead of a document. A school committee 20 miles outside Boston already worked out which one actually works.
More Resources:
- [AI Isn’t Killing Marketing Accountability, It’s Exposing Who Never Had It](https://www.searchenginejournal.com/ai-isnt-killing-marketing-accountability-its-exposing-who-never-had-it/586486/)
- [Data Shows AI Citation Patterns Reveal Strategic SEO Opportunities](https://www.searchenginejournal.com/comparison-of-ai-citation-patterns-offers-strategic-seo-insights/573327/)
- [Only 28% Of Americans Trust AI Search – And That Gap Is Your SEO Opening](https://www.searchenginejournal.com/only-28-of-americans-trust-ai-search-and-that-gap-is-your-seo-opening/581913/)
Featured Image: Alonchik_73/Shutterstock