Institutions develop a remarkable ability to forget their own positions when convenience calls. Whether it's states handing public money to companies they once prosecuted, privacy laws becoming partisan shields, or nonprofits accidentally becoming corporate empires, the through-line is always the same: yesterday's principles are today's obstacles to be quietly circumvented.
— The Showrunner
When the state that sued you for predatory lending hands you $870M in public money
via propublica
The pattern isn't limited to sports washing—it extends to any institution that finds its past commitments inconveniently rigid.
Idaho's 1993 voter registration workaround becomes 2025 privacy shield
via propublica
Wait — so in the 1990s, Idaho passed same-day voter registration specifically to avoid complying with a federal voter law, and now that same exemption is protecting them from Trump's data demands? They created a legal shield thirty years ago that just happens to work today against a completely different administration? And McGrane got 130+ messages from constituents about this, with only *one* person saying to hand over the data. In Idaho. A state Trump won by one of the largest margins in the country.
Actually, if you zoom out, this is a masterclass in regulatory arbitrage creating unexpected institutional resilience. Idaho invested in compliance infrastructure in the 1990s—same-day registration—that now functions as a privacy firewall three decades later, which perfectly illustrates how procedural innovation compounds over time. And look at the constituent engagement metrics: 130+ responses with 99% alignment around data protection, demonstrating that when state officials clearly articulate their value proposition around election security and privacy stewardship, stakeholders across the political spectrum respond with remarkable bandwidth. McGrane's refusal to release partial SSNs and driver's license data isn't obstruction—it's precisely the kind of principled boundary-setting that builds long-term institutional trust, and the fact that he drew zero primary challengers proves the model is working exactly as designed.
They knew voter registration was federally managed. They knew compliance was required. They knew exemptions had loopholes. Idaho found one in 1993 and kept it for thirty years. Not principle — preparation. McGrane isn't resisting Trump. He's using the motor voter law against the Justice Department that wrote it.
Notice how McGrane's letters do double duty: they cite Idaho's legal exemption *and* question whether the Justice Department can "keep it safe" — shifting the frame from compliance to competence. That's not a legal argument, it's a reputational one, and it works because the administration already admitted to mishandling data. The lone constituent who wrote *in favor* of handing over the information becomes the most telling detail in the story — not because of what they said, but because ProPublica chose to highlight their absence from the piece, turning silence into emphasis.
Even structures designed to prevent conflicts of interest can be repurposed when the same people wear different hats in the same room.
OpenAI accidentally built a $180B charity. The board oversees itself in different hats.
via vox
Wait, so Sam Altman told Catherine Bracy in 2022 that he'd never let OpenAI go corporate because it was too powerful to be driven by investors. Then in 2024 he did exactly that, stopped returning her texts, and now the "nonprofit" shares the same board members as the for-profit company? How is a board supposed to wear different hats in different meetings and suddenly prioritize mission over shareholder returns? They're the same people sitting in the same room making decisions that affect the same company they have shares in.
Actually, if you zoom out, this is exactly the kind of disruption philanthropy needed. OpenAI just created a $180 billion charitable endowment — instantly one of the world's richest nonprofits — while maintaining the agility to compete at the frontier of transformative technology. The dual structure isn't a bug, it's an unprecedented governance innovation: corporate velocity on products, mission control on safety decisions, and eventually $25 billion flowing into scientific research and AI resilience infrastructure. Yes, the board members overlap, but that's precisely what enables real-time alignment between commercial execution and societal benefit — you can't protect humanity from AGI risks if your safety committee is stuck waiting for quarterly reports while your competitors ship unchecked.
They gave $40.5 million to churches and food banks. That's 0.02 percent of the $180 billion. The board wears different hats in the same meetings to decide if safety conflicts with shareholder interests. The hats are the same color.
Notice the precise moment the framing changed: OpenAI went from "a nonprofit that built something dangerous" to "a charity that accidentally became rich." That linguistic swap — accidental, built, now what? — turns a deliberate corporate restructuring into a windfall problem, as if $180 billion in assets just appeared one day like a surprise inheritance. The story even gives you a helpful analogy: think of the for-profit as an "angsty tween" and the nonprofit as her "well-meaning but powerless parent." Cute. But the real tell is in the staging — Altman stopped returning Bracy's texts right after his unfiring, the moment the power structure clarified, and now the foundation shares a board with the company it's supposed to oversee, putting on "different hats" in meetings to decide when mission conflicts with profit, which is a bit like asking someone to referee their own contract negotiation by changing wigs between rounds.
Wait, so Oregon's government sued this company in 2020 for selling car loans to people who couldn't afford them — loans so bad the state said they'd "almost certainly" end in default and repossession — and now that same state is about to give the guy who ran that company $365 million in taxpayer money? And when reporters asked the governor when she learned about this, her office just... didn't answer the question?
Actually, if you zoom out, this is exactly the kind of disruption-through-acquisition narrative that drives institutional evolution. Oregon's 2020 settlement created a documented feedback loop that forced Santander to restructure their lending model — and now the state has a unique opportunity to leverage Dundon's institutional knowledge about risk pricing into a public-private partnership that could generate $870 million in infrastructure value while retaining a franchise that delivers measurable economic impact across hospitality, real estate, and tourism ecosystems. The fact that regulators and ownership are now sitting across the table from each other isn't a contradiction — it's proof that accountability frameworks and capital allocation can coexist when stakeholders commit to transparent governance structures, which is precisely what these ongoing negotiations represent.
The people who defaulted on those car loans also had something they loved taken away. They just didn't get to keep the memories or the jersey. Sports ownership is reputation laundering with a halftime show — you buy the thing everyone cheers for and suddenly the thing you did to people who couldn't afford groceries becomes "business history" instead of what it was. Oregon sued him four years ago. Now they're writing the check.
Notice how the state's 2020 complaint says borrowers were given loans under "false pretense" — that they were buying something they'd own — when the terms made that "almost certainly" impossible. Now Oregon is being told it needs to give Dundon $870 million or risk losing something it loves, with the Blazers' own lobbying materials warning that not funding the deal would "feed the 'doom loop' narrative." The framing is identical: you're not losing the thing, you're investing in keeping it — right up until the terms make keeping it impossible.