Fragment 1/8

English draft

Opening

Reader, Maysville, Kentucky is a small town of 8,700 people. Last year, an unfamiliar company approached local farmers there, offering to buy their land. The price on the table was roughly 10 times market value.

The offer made to Delcia Bear and her mother, Ida Huddleston, came to $26.48 million combined — about ₩39 billion in Korean currency terms. The two women signed at first. Then, a few months later, they reversed the deal.

Stories like this usually end with some moral about “what matters more than money.” But I see something different in this case. Let me give you the conclusion up front: what overturned this deal wasn’t the amount — it was information. And the very thing that blocked that information was the same tool the buyer had reached for to move fast.

---

A $26.48 Million Offer Lands

Let’s start with the numbers. The offer was to buy 463 acres of Bear’s farm at $48,000 per acre, and 71 acres of his mother’s land at $60,000 per acre. Combined, the two parcels come to $26.48 million.

To get a sense of what that means in Maysville terms: the median household income in this area is about $39,000. That’s less than half the U.S. average, and nearly 40% below Kentucky’s state average. One in four residents lives below the poverty line. In other words, this is the kind of money that changes a life completely.

The reason someone wanted this land was a hyperscale data center¹ spanning 2,080 acres. Its power draw would be 2.2 gigawatts — roughly what 1.65 million households use at once. Picture the electricity demand of several mid-sized cities dropping in next to a town of 8,700 people.

There were genuinely attractive terms on the table for the region, too. 400 full-time jobs, plus hundreds of construction jobs. With the county’s working-age population at around 11,000, that’s not a small number. Projections suggested property and facility taxes could double or triple the city’s budget. There were commitments of at least $15 million for police, fire, and EMS, and $80 million for water system upgrades. For context, the water pipes in this town burst about once a month.

Tyler McHugh, the local economic development official, summed up public opinion this way: 20% strongly opposed, 20% strongly in favor, and the remaining 60% just wanting the whole debate to be over. I’d argue that 60% is the real protagonist of this story.

---

The One Line on Page Seven of the Contract

This is where the real design of the deal comes into view.

At the top of the seventh page of the purchase agreement was a clause. It stated that no information related to the deal, whether spoken or written, could be disclosed to any other individual or organization. Mother Huddleston reportedly got angry at this passage. And she did exactly the opposite. She called her neighbors together and started comparing the offers each of them had received.

Here’s why this scene is pivotal: up to that point, the sellers had no basis whatsoever for judging whether their terms were good or bad. The figure of “10 times market value” was information the buyer’s side had provided. They had no idea what the neighbors were getting, what would be built on the land, or who the buyer actually was. All the development agent could reveal was a single line: “a Fortune 500 tech company.”

In economics, this situation is called information asymmetry². When only one side knows the true nature of what’s being transacted, the market stops being about price and becomes a matter of trust. And in a market where trust has collapsed, even good terms get rejected.

That’s exactly what happened. In July 2025, residents learned that the project was a massive data center. Bear succeeded in unwinding both contracts, arguing that she and her mother had signed without full knowledge of the buyer’s plans. What’s interesting is how the buyer’s side reacted. According to McHugh, the company said it would have terminated the contracts anyway—their reasoning being that having a reluctant seller as a neighbor was not a good start.

There’s something worth noting fairly here. The non-disclosure agreement isn’t a malicious device. The moment a candidate site becomes known, land prices spike, competitors move in on the same area, and negotiating leverage evaporates. For a company comparing multiple sites, it’s a rational tool. In fact, it’s standard practice in the data center industry, and McHugh himself signed one, putting himself in a position where he couldn’t say what he knew.

The problem lies in the premise the tool was designed around. Non-disclosure agreements are built on the assumption that the counterparty is a corporation. But here, the counterparty was a community of neighbors who had lived in the same place for nearly 200 years. In a place where knowing each other’s business is the default, cutting off the flow of information isn’t a negotiating tool—it’s a relationship-destruction device.

---

The Second Ledger That Secrecy Erased

Blocking information hides one more thing: the fact that benefits and costs land on different people.

The person who sold the land gets a check. The person who didn’t sell gets nothing—and gets 2.2 gigawatts next door. Behr himself admits the irony. His refusal doesn’t stop the data center from getting built. In fact, the company simply redesigned the layout around the landowners who agreed to sell. That leaves an open-ended outcome in which the person who said no bears the burden with no compensation at all.

This structure produced a lawsuit. Opponents filed two suits challenging the ordinance and the rezoning³, but under Kentucky legal procedure, they had to name as defendants the very landowners who had consented to the rezoning. People who had never set foot in a courtroom in their lives found themselves served with a complaint by a neighbor living 3 kilometers away. The judge assigned to the first case was found to have a conflict of interest—his spouse was the county clerk who took the meeting minutes, and his nephew was active in an online group supporting the project—so the case was transferred to another jurisdiction. In the wake of the litigation, the company delayed payment for the land sales, which meant other farm owners still waiting on their money got caught up in the mess too.

On the flip side, there’s a case where disclosure immediately created bargaining power. The 28 households in a mobile home park located inside the project site were initially notified of a $20,000 relocation payment. Janet Garrison, a former local IT instructor, contacted a news outlet and arranged interviews with elderly residents; within days, the relocation payment rose to $50,000. Garrison later ran for county commissioner on a platform opposing the project and lost by 75 votes out of 1,631 cast.

Here’s the summary: secrecy protected the buyer’s bargaining power, but at the same time it stripped the community of its ability to reach its own consensus. When no one knows who’s getting how much, collective action becomes impossible, and all that’s left is everyone fending for themselves amid mutual distrust.

For what it’s worth, the buyer’s identity still hasn’t been officially confirmed. A public-records request turned up circumstantial evidence that the registered agent of the corporation set up for the project had been involved in several other projects later revealed to be Meta sites—but Meta’s position is that it has made no decision to enter that region. I think it’s premature to name the company at this stage. Still, I want to flag that the very fact that it can’t be named is the crux of today’s story.

---

What’s missing in Korea isn’t information—it’s procedure

There’s a reason this story is hard to dismiss as someone else’s problem.

In 2025, at least 48 data center projects across the United States were halted or delayed by resident pushback. But Korea’s numbers are worse. According to a recent tally, of 33 data center permits in the Seoul metropolitan area, 17 were delayed or scrapped due to resident opposition. That’s more than half. In Doksan-dong, Geumcheon District, Seoul, construction was fully halted and the case went as far as a lawsuit to revoke the building permit. In Yongin and Anyang, Gyeonggi Province, plans were withdrawn. Goyang saw a roughly 10-month delay, and Ichon-dong, Yongsan District, Seoul, is still stuck in the process.

Here’s where an interesting difference emerges. In Korea, who the developer is tends to be public knowledge. Kentucky-style anonymous acquisition through shell buyers is rare. And yet the outcomes are just as bad, or worse.

So what’s missing in Korea isn’t information about the buyer. It’s the procedure that determines when residents get to weigh in, on what standing, and about what. A resident briefing session held after the permit is effectively already locked in isn’t a briefing—it’s a notification. And the only tools left to people who’ve just been notified are banners and lawsuits.

In the American cases, what opposing attorneys kept demanding wasn’t that the project be stopped—it was a Community Benefits Agreement⁴. If you can’t stop the company from taking its profit, the argument goes, at least put in writing what the community gets in return. Korea has essentially no equivalent mechanism. Instead, compensation comes as donations or funding for local pet projects, handled case by case. Individual payouts split what is really a shared cost into a handful of personal checks—a structure that is practically guaranteed to fracture communities from within. It’s the same path that turned neighbors in Maysville into plaintiffs against each other.

Korea’s private data center market is projected to grow from roughly ₩6.22 trillion in 2024 to about ₩10.19 trillion by 2028. If the procedure isn’t designed now, a meaningful share of that growth will simply burn away in the permitting waiting room.

---

Oswald’s Lens

A few years ago I wrote a paper on how information asymmetry hardens into exploitative business structures. One conclusion from that work has stuck with me. An information gap does its worst damage not when you deceive the other side, but when you deny them the material they’d need to judge for themselves. The Kentucky case is exactly that structure. Nobody lied. They just made sure no one could say anything.

I’ve watched this scene play out repeatedly while building corporate strategy. A company treats site acquisition as a procurement problem: lowest price, shortest timeline, minimum exposure. From a procurement standpoint, an NDA is the right answer. But what’s actually happening isn’t procurement — it’s market entry. You’re forming a relationship in which you’ll share water and electricity with the same neighborhood for the next 20 years. Seen through the lens of entry, opening the first conversation with control is the worst possible move.

This is where the paradox of speed shows up. NDAs exist to save 6 months. But in Maysville, trying to save those 6 months bought 2 lawsuits, delayed payments, and a court calendar with no end in sight. I have a rule of thumb I use in consulting: if the time you lose when a shortcut fails is more than 3 times the time you saved, it wasn’t a shortcut — it was a bet. In this case, it was clearly a bet.

And this isn’t just a data-center story. The same thing happens inside organizations. Whether it’s a business-unit reorganization or a change to personnel policy, consent given by people who don’t have the full picture always gets revoked later. Not because the signature itself is invalid, but because trust evaporates the moment the missing information finally arrives. I’ve seen organizational-change projects treat “when do we tell them” as a communications problem, over and over — but it isn’t a communications problem. It’s a question of whether consent is valid at all.

---

Closing

Let me sum this up in three lines.

First, the reason the 39 billion won offer was rejected wasn’t the amount — it was the information. The sellers signed without knowing what they were agreeing to, and once they found out, they reversed course.

Second, the non-disclosure agreement protected the buyer’s negotiating leverage, but it stripped the community of its own capacity to build consensus. The result was lawsuits between neighbors and delayed payments.

Third, Korea is living through the same outcome even when residents know who the developer is. What’s missing here isn’t information — it’s a procedure that defines when and in what capacity residents can intervene. 17 of 33 permit cases in the Seoul metropolitan area are stuck in that exact gap.

If you’re preparing a major change inside your organization, I’d suggest asking this question before you set the announcement date: when the information we’re currently withholding reaches the other side, will this agreement still hold? If it doesn’t feel like it will, then what you have isn’t an agreement yet — it’s a postponement.

How about you, Reader? If you’ve ever had a moment where, after learning the full picture later, you thought “if I’d known this, I never would have agreed” — tell me in the comments what piece of information was missing at the time. If enough stories come in, I’ll pull them together in a future issue as a pattern of information asymmetry inside organizations.

---

💬 Tell me in the comments about a moment when you regretted agreeing to something after learning the full picture later — and what information was missing at the time. I’ll factor it into a future issue. 📨 If you have a colleague who works in permitting or local negotiations, please pass this piece along to them.

---

Found this useful? Share it

Copy linkXLinkedInThreadsKakaoTalk네이버 밴드

References & Further Reading

Primary sources

• Joseph Pisani, “The ‘Country Hicks’ Who Refused $26 Million from an AI Data Center”, The Wall Street Journal, 2026. ··· This is the backbone of today’s piece. Pay special attention to the page-7 non-disclosure clause in the contract.
• Ledger Independent, “Second complaint filed amid proposed data center project in Mason Co.”, July 2026. ··· The plaintiff and defendant lists appear in full here — you can see the structure that forced neighbors to name neighbors as defendants.
• WEKU, “Mason County gives final approval on rezoning properties for data center construction”, May 26, 2026. ··· Contains the opposing counsel’s remarks demanding a Community Benefits Agreement.
• Newspim, “Will the Yongsan data center find a solution at the August 6 resident briefing?”, Aug 6, 2026. ··· The source for the figure that 17 of 33 permit applications in the greater Seoul area were delayed or scrapped. This is the basis for the Korea section.

Background

• WVXU, “Data center near Maysville sparks debate about land use”, Aug 4, 2026. ··· The most detailed local reporting on the situation facing mobile-home-park residents.
• Climate & Energy Economy (Kihu Energy Kyungjae), “Data centers devouring power and water: 48 US projects derailed by resident backlash”, 2026. ··· A tally of stalled projects nationwide — it shows Kentucky is not the exception.
• Electric Times (Jeongi Sinmun), “The AI power war has begun”, Jan 15, 2026. ··· A feature that examines both the outlook for Korea’s domestic data center market and the grid bottleneck.

Past issues worth pairing with this one

• Why Google Chose a Reactor It Had Never Built Before ··· If today’s piece is about the siting problem for data centers, this one is about the power problem. Read them side by side and the full picture emerges.
• What the Luddites Smashed Wasn’t Machines ··· Connects on the axis of resistance to the terms of technology adoption. Worth reading as the 200-year-old version of this story.

---

The author is Oswarld (Kwangseob Ahn). Current roles: Adjunct Professor at Sejong University, Strategy Consultant at INLEVEL9. Career, research, books, and recent work are kept current on the About page. Latest · July 2026: HEMA-2: A Consolidation-Aware Tri-Memory Architecture with Multi-Channel Scheduling for Lifelong Conversational AI.

📝 Glossary

Footnotes

1. Hyperscale data center: A massive facility that operates hundreds of thousands of servers on a single site. While an ordinary data center draws 10–25 megawatts, these facilities routinely exceed 100 megawatts — from the grid’s perspective, it’s like a new city springing up overnight. ↩

2. Information asymmetry: A situation where one party to a transaction holds critical information the other lacks. The classic example is the used-car market, where only the seller knows the car’s defects — leaving the buyer to judge not by price but by suspicion. ↩

3. Rezoning: The administrative process of changing the designated use of a specific piece of land — for instance, converting farmland into industrial land. It requires approval from local planning commissions and councils, which is exactly where resident opposition actually collides with the process. ↩

4. Community Benefits Agreement: A binding contract that a large developer signs with a local community. It locks in commitments — hiring quotas, infrastructure investment, noise limits — in writing, so that verbal promises can’t later be walked back. ↩