James Freeman’s June 3 “Best of the Web” column is a donor-class permission slip — a tightly engineered piece of apparatus prose designed to train the Journal’s business-conservative readers to see any government involvement in artificial intelligence as an attack on their freedom, their prosperity, and their country. We drafted these deregulatory permission slips for two decades, translating balance-sheet requirements into patriotic appeals to keep the regulatory gap open long enough for the client to cash out. And watch the timing: the column drops the morning after Trump signs a downsized AI executive order, less than two weeks after the president himself killed an earlier draft because he worried it might slow U.S. advantage. Freeman’s “sadly the president is now suddenly looking less zealous” is a preemptive strike on the impulse that survived the purge — and a clinic in the old art.

This column recently lauded President Donald Trump’s deregulatory zeal and warned that his extremely wise decision to reject Biden policy and endorse the freedom of Americans to develop artificial-intelligence technologies was in danger of being reversed. Sadly the president is now suddenly looking less zealous.

Frame-engineered relabeling — WSJ §A.1 — operates through the substitution of “deregulatory zeal” for regulatory capture, and “freedom” for corporate immunity. The operator’s-eye-view is straightforward: the writer establishes a moral baseline in which unregulated corporate deployment is the good and the state’s basic interest in model security is a “reversal.” This is the same semantic architecture that built “tax relief.” “Zeal” pathologizes caution; “freedom” sanctifies the balance sheet. The actual order, as reported, asks companies to give the government early access to powerful AI models for security review and explicitly bars “a mandatory governmental licensing, preclearance, or permitting requirement.” That is not a costly mandate; it is a voluntary security conversation. But the column never tells the reader that. Instead, the bait-and-switch installs the frame before any facts appear.

That was the right instinct. Unfortunately the new compromise poses a risk to U.S. technological vitality and prosperity. Mr. Ramkumar mentions Dean Ball, a former Trump adviser: This is exactly the risk in this market where the U.S. is the leader and which promises enormous potential to improve productivity and raise living standards… What are the companies supposed to do when government officials respond to each new model with a list of bureaucratic suggestions allegedly intended to improve the software?

Multiple-audience targeting — WSJ §A.3 — runs at full throttle. The populist base gets the anti-Washington grievance (“costly mandates,” “never explicitly approved by Congress”). The technocrat gets the policy-wonk frame. The donor-class reader gets the permission slip: the government is an incompetent bureaucracy inventing hurdles out of thin air. We wrote columns like this to make the safety engineer look like a petty tyrant and the CEO look like the innovator. Meanwhile the rhetorical question at the end is the strawman’s delivery vehicle — “bureaucratic suggestions allegedly intended to improve the software” treats even a voluntary safety query as a prelude to tyranny. The column trains its audience to recoil from the lightest oversight because “bureaucratic” is always bad and “suggestions” are always a threat. The order’s own words are irrelevant; the alarm is already tripped.

Yet the poll also asked participants to respond to the following statement: A full 57% of registered voters in the survey agree, while only 30% disagree. But these positive changes will not happen if AI is directed by the federal government.

The column omits the exact poll statement, then cherry-picks a single 57% figure from the same Economist/YouGov cycle that found a broadly sour public mood on AI — and jumps to the conclusion that the hopeful minority would never get what it wants if the government is involved. Selective attention — Bad-Faith Catalog: selective_attention — executes the pivot. The 57% agreed the technology holds promise; they did not agree that federal safety inspectors should be banned from the engine room. The column conflates the two to manufacture a policy mandate. In the layout game, we used this move constantly: take the sliver of public hope left after years of trust-eroding spin, and wield it against the very guardrails the public needs. It is the gaslight register, polished to an op-ed shine.

Alex Tabarrok writes at Marginal Revolution with a reminder of the importance of America’s services economy and the possibilities for an AI export boom: He quotes economist Richard Baldwin, who writes: Are We Negotiating with an Ayatollah or an AI-atollah? Given the fact that Iran’s new Ayatollah Mojtaba Khamenei has not been seen in public…

Authority laundering — Bandura: displacement of responsibility — drapes an Iran tangent over the deregulatory argument like a flag on a coffin. Tabarrok and Baldwin supply the intellectual cover, then the column drives straight into a foreign-policy threat designed to spike the reader’s cortisol. The “AI-atollah” pun is pure threat-inflation closer — WSJ §A.13 — and it is meant to keep the audience in threat-salience while the donor class enjoys the domestic security vacuum. The reader’s pulse quickens; the balance sheet fades.

Catholic archbishop of New York Ronald Hicks writes in First Things on the rise of physician-assisted suicide: [Hicks].

The final pivot to Archbishop Hicks is brochure-fragment architecture — a jolt of existential dread meant to leave the reader with diffuse moral unease rather than a policy question. The piece is not a coherent argument; it is an emotional carousel, moving the reader from deregulatory panic to foreign-policy anxiety to end-of-life morality without ever letting them look directly at the balance sheet. Every stop supplies a different fear; every stop hides the same donor interest.

Here is what the Freeman artifact actually amounts to, taken together. The donor class needs the regulatory gap to hold long enough to secure its AI margins, and the Journal’s job is to translate that balance-sheet requirement into a patriotic emergency. The reader gets to feel like a patriot protecting American innovation while the apparatus transfers the risk of the next systemic failure onto the same reader’s shoulders. The con is old enough to vote, and it works because it flatters the reader: you are not a consumer who might want a say in how this technology shapes your life; you are a defender of freedom, and anyone who questions the industry’s right to move fast and break things is a bureaucrat with a sad history. The output is not confusion; it is a tripped alarm system. The reader is now calibrated to treat a security phone-call as a hostile takeover. The column is a grift wrapped in a flag, and the only thing entirely artificial about it is the intelligence that pretends the balance sheet isn’t the only page that matters. We know because we helped write it.

— Phukher Tarlson