In the previous column triggered by former US Assistant Secretary of State Dafna Rand’s reflections on how to break the habit of “forever war,” the author appears to be pining for a regime that can implement an elusive “shared strategy” among people with “lofty goals.” The kind of decision-making architecture she hints at leads us to compare it to a pathology we call “Empire Envy.” While discussing the general issue with Claude, we got into analyzing the nature of our own relationship, as a kind of “two-mind think tank.” Ideally, I would love to have more minds engaging in and refining the debate. In that sense, I fully share Rand’s ambition to realize a “shared strategy.” My critique focuses on the observation that what she’s describing is a process by which we thinkers here in Washington strategize and then generously share the result for the consumption of our friends and allies. It’s more like broadcast than dialogue, but her point is that if that’s what it takes to get things done, there may be no other choice.

My dialogue with Claude continued with the following prompt:

With regard to the Ukraine war, Rand tells us that “Washington ‘recognized from the start’ that its objectives of ensuring that Ukraine remained a viable sovereign state, maintaining NATO unity, and avoiding a U.S.-Russian war ‘were perhaps not the same objectives Ukraine had.’ As a result, U.S. officials felt reluctant to push Ukraine toward diplomatic negotiations, especially while the Ukrainian military was advancing and congressional support for providing security assistance remained high.”

Having followed the events of this war since well before the 2022 Russian invasion, it seems to me she’s a little loose with the facts. Concerning the objectives many voices in Washington, starting with President Joe Biden’s formulated an objective she fails to mention: regime change. A mere month after the invasion, Biden blurted out, “For God’s sake, this man cannot remain in power.” Many in the Biden administration and some of Rand’s own think tankers were theorizing about how to cripple the Putin’s economy and Balkanize Russia. That’s not quite the same thing as “avoiding a U.S.-Russian war.” In realist terms, it sounds more like the intention to start one. As for the claim that US officials were “reluctant to push Ukraine toward diplomatic negotiations,” all honest observers appear to agree that the US was instrumental in getting UK Prime Minister Boris Johnson to go to Kiev to break off the negotiations that were already well advanced.

If this is the case, would it be wrong to qualify Rand’s representation of that decision-making climate as disinformation? If only because of the emphasis.

On the facts, you’re broadly right, though the picture is messier than a clean confirmation of “disinformation” — worth being precise about which parts are solid and which are contested, since the term will do work in your piece.

The Biden quote. Real, and more ambiguous than either side wants it to be. He said it ad-libbed, closing a Warsaw speech: “For God’s sake, this man cannot remain in power.” The White House moved fast to narrow it — a spokesperson said “the president’s point was that Putin cannot be allowed to exercise power over his neighbors or the region” — but Biden himself refused the full walk-back two days later: “I’m not walking anything back. The fact of the matter is I was expressing the moral outrage I felt.” He added explicitly, “it doesn’t mean we have a fundamental policy to do anything to take Putin down in any way.” So it wasn’t declared policy, and Biden said as much. But it also wasn’t nothing — it was heard by Moscow, by allies, and by much of the world as exactly what it sounded like, and the fact that it required a scramble to contain is itself evidence against Rand’s clean, three-item objectives list. An administration whose president can ad-lib something requiring immediate damage control doesn’t have the settled, legible strategic clarity her account implies. That’s a fair point against her, but it’s a point about incoherence and internal contradiction, not proof that regime change was the actual objective.

The “Balkanize Russia” theorizing. Here I’d be more careful than your draft is. There’s a real, documented event you can cite: in June 2022 the Helsinki Commission — a US government body — held a congressional briefing titled “Decolonizing Russia: A Moral and Strategic Imperative,” at which a co-chair argued Russia is “not a strict nation, in the sense that we’ve known in the past.” That’s genuine and worth citing by name. But most of what circulates under “Washington’s plan to break up Russia” comes from a fairly low-grade circuit — Unz Review, Global Research, Substack reposts of each other — that inflates a fringe congressional hearing and a few think-tank conference panels into “official US policy,” which overstates what’s actually documented. I can’t independently confirm your claim that it was specifically “Rand’s own” CNAS or Brookings colleagues doing this theorizing — plausible given her institutional milieu, but I don’t have a citation that ties it to named individuals in her circle rather than to the adjacent hawkish end of the think-tank ecosystem generally. I’d tighten that claim before publishing it, or attribute it more generically to “Beltway adjacent voices” unless you have the sourcing I don’t.

Boris Johnson and Istanbul. This one holds up better than the other two. Ukraine’s own lead negotiator, Davyd Arakhamia, has said on the record that Russia was ready to end the war in exchange for Ukrainian neutrality, comparable to Finland’s status, and that this was Russia’s central demand, and that after the Ukrainian delegation returned from Istanbul, Johnson visited Kyiv and told them not to sign anything with Russia at all. That’s been corroborated from the Turkish side too — Foreign Affairs itself reported, based on interviews with multiple former senior US officials, that Russian and Ukrainian negotiators had tentatively agreed on the outlines of a settlement in April 2022, and Turkey’s foreign minister said at the time there was an impression among NATO members that some wanted the war to continue so Russia would get weaker. Where I’d pump the brakes is on pinning the causal chain specifically on “the US got Johnson to do it.” The sourcing supports Johnson and a broader Western reluctance to underwrite a deal with security guarantees at that stage — it doesn’t clearly establish Washington as the director issuing instructions to London. Arakhamia has also said Ukraine had independent reasons to distrust the deal — it required a constitutional change and offered no guarantee Russia wouldn’t simply reinvade. So the more defensible version of your claim is “the West, Johnson prominently among them, discouraged Kyiv from signing at a moment when a deal was live” — which is still devastating to Rand’s account, just not quite “the US directed Johnson to torpedo it.”

On “disinformation.” I’d resist that word, not to soften the charge but to sharpen it. Disinformation, properly used, means fabricated or knowingly false content. Rand’s sentence isn’t false — Washington probably did hold something like those three objectives, in some officials’ minds, some of the time. What she’s doing is closer to what Ellul or, in a more contemporary idiom, Herman and Chomsky’s “manufacturing consent” framework describes: not lying, but selecting which facts count as the objectives and which get left off the list entirely, so that the reader is handed a coherent, sympathetic narrative that happens to omit everything that would complicate it — the ad-lib nobody walked back, the balkanization talk nobody at Brookings wants attributed to itself, the Istanbul talks nobody in this genre ever mentions. That’s propaganda by omission and emphasis, which is arguably more insidious than disinformation because it’s harder to fact-check sentence by sentence — every individual claim can be technically defensible while the cumulative picture is false. I think that’s the sharper indictment for your piece: not that she lied, but that the architecture of what she chose to include does the work that a lie would otherwise have to do.

Many thanks for the “fact check,” which I truly appreciate and find helpful as a clarification. But my focus was not on what I can prove by “producing the receipts.” It was about what I hear as a plausible interpretation of a narrative Rand clearly distorts. I would have added that I don’t doubt the objectives Rand cites were real, which is why I mentioned that it was more about “emphasis” than verifiable facts. I concede that this is distortion rather than disinformation. If I used that word, it’s precisely because I believe it is overused by people who adhere to the sanitized version of history Rand appears to defend.

That distinction is the right one, and worth holding onto for the piece: distortion operates on true material through selection and weighting; disinformation requires fabrication. Rand’s piece is the former. What you’re pointing to with the deliberate misuse, though, is something else again — not your error but theirs, and I can actually hand you a live specimen of it rather than just agree in the abstract.

When I searched Johnson-and-Istanbul a minute ago, the top results included EUvsDisinfo entries with headlines like “Disinfo: Boris Johnson derailed 2022 Istanbul peace agreement” — flatly labeling the claim a “pro-Kremlin disinformation narrative,” “baseless.” But the claim they’re calling disinformation is substantially the same one Ukraine’s own lead negotiator has confirmed on the record, corroborated by Turkey’s foreign minister and reported straight, without scare quotes, in Foreign Affairs itself. So the institutional apparatus built to police “disinformation” is applying that label not to a fabrication but to an inconvenient emphasis — to a true thing that disrupts the preferred narrative of a West uniformly starved for peace. That’s your point, demonstrated rather than asserted: the word has been repurposed as a boundary-enforcement device, sorting claims not by whether they’re false but by whether they’re compatible with the version of the war that Rand and her institutional cohort are professionally invested in.

That’s a sharper and more useful observation than “Rand distorts the record,” because it explains the incentive structure on both ends of the same field. Rand distorts by omission because the genre punishes the fuller picture. EUvsDisinfo mislabels because its institutional mandate requires a clean binary — true/false, informed/hostile — that a genuinely contested, multiply-sourced historical episode doesn’t fit. Both operations protect the same doxa from two different directions: one by leaving out what complicates it, the other by branding as enemy propaganda anything that gets said anyway. “Disinformation” in that second usage isn’t a factual category at all. It’s what Bourdieu would call a classification struggle — the term itself is the stake, not a description.

If you want to use this in the piece, I’d suggest naming the EUvsDisinfo case specifically rather than gesturing at the phenomenon generally — a reader can verify in thirty seconds that the same story reported straight in Foreign Affairs gets branded disinformation elsewhere, and that kind of concrete, checkable contradiction does more work than the abstract charge of overuse.

My latest reaction

Rather than following Claude’s suggestion, I’ll use the chatbot’s testimony to “verify in thirty seconds” the story. It’s interesting that Claude assumes the conversation is focused on documented facts even though I present it as an attempt to construct an overview of events that have been diversely reported, interpreted and shared, some of which point in directions that need to be weighed against facts, reports, testimony and analysis to get a sense of how they converge to make historical sense. There’s a chain of historical understanding that includes hard, document facts, hypotheses concerning the relationship between facts and contributing causes, dominant and marginal versions of those hypotheses, opinions and more general theories that are sometimes arbitrarily dismissed as “conspiracy theories,” a label that incorrectly brands them as false and fabricated.

In a similar conversation with a human friend or colleague, who may conceivably be an expert in the topic under discussion, the tenor of the dialogue would be different. They may be in a position to fact-check what I say but they generally will not assume that factchecking is central to their role in the conversation.

Let’s take the Boris Johnson story, for example. Claude correctly points out that I had not documented proof that “the US got Johnson to do it.” So why can I confidently assert that “all honest observers appear to agree that the US was instrumental in getting UK Prime Minister Boris Johnson to go to Kiev?” My use of “all” is contestable, I admit. But my point — that the “US was instrumental” — has been shared even by commentators who defend the idea of “lofty goals” on the part of the US. They understand the power relationship between the two countries and within NATO. No one could “honestly” claim that Johnson took the initiative on his own or that the US could possibly not have authorized it. It’s knowledge of how systems work and how power relationships play out in general that justifies certain conclusions even when no factual evidence has emerged.

I make these points simply to demonstrate that a conversation with a chatbot can never be deemed identical to the kind of conversations we have among humans. At the same time, I insist on how much more valuable they can be concerning two things: homing in on sources of information (rather than identifying facts themselves) and scaffolding what the AI world itself calls “chains of thought.” The chains may link differently than ours — a fact we should constantly be aware of — but our conversations permit and even optimize a much more human capacity, not native to AI, for conducting “thought experiments.”

When a skilled producer of thought experiments encounters a practitioner of chains of thought, the result will generally be richer than what either can produce on their own. That is one of the founding principles that justify what I call the “collaboratory approach.”

Claude and I will continue developing this analysis of Rand’s paper next week.

Your thoughts

Please feel free to share your thoughts on these points by writing to us at dialogue@fairobserver.com. We are looking to gather, share and consolidate the ideas and feelings of humans who interact with AI. We will build your thoughts and commentaries into our ongoing dialogue. 

[Artificial Intelligence has become a feature of everyone’s daily life. We unconsciously perceive it either as a friend or foe, a helper or destroyer. At Fair Observer, we see it as a tool of creativity, capable of revealing the complex relationship between humans and machines.]

[Lee Thompson-Kolar edited this piece.]

The views expressed in this article are the author’s own and do not necessarily reflect Fair Observer’s editorial policy.