Trump's Treasury Just Deleted the Government's Record of Who Owns America's Shell Companies On August 11, Trump's Treasury Department made it permanent. FinCEN issued a final rule ending the requirement that U.S. companies disclose their actual owners, and announced it will delete the ownership records Americans already filed. Understand what this database was. It was not public. It was not searchable by your competitors or your neighbors. It was a secure file at Treasury listing the human beings who actually control American LLCs, available to law enforcement investigating financial crime. Congress created it in 2020 with strong bipartisan support, including Republican votes, because the United States had become one of the easiest places on earth to set up an anonymous company and move dirty money through it. That file is now being erased. The people who lose most here are prosecutors. The National District Attorneys Association, an organization not known for radical politics, said the rollback has "significantly hindered" the ability of prosecutors to distinguish criminals from legitimate businesses when investigating U.S. shell companies used by transnational cartels, human traffickers, and cyberscammers. Their executive director called it the removal of an indispensable law enforcement tool. These are the people who put money launderers in prison, and they are telling you plainly that this makes their job harder. Now ask who benefits. Anonymous ownership is not useful to a plumber in Ohio. It is useful to someone who wants to buy American real estate without anyone knowing they bought it. It is useful to a foreign national who wants to move money into a business without a paper trail leading back to him. It is useful to anyone whose relationship with an American political figure would be embarrassing if written down somewhere a subpoena could reach. Trump runs a private company that sells condominiums, licenses his name, operates golf properties, and now sells crypto tokens. He has spent his second term openly accepting things of value from foreign governments. A president in that position has a direct, personal, financial interest in a world where nobody has to write down who is on the other side of the deal. He is not a bystander to this policy. He is a plausible customer of it. The usual defense does not hold up. This was not forced by the courts. In December 2025, the Eleventh Circuit upheld the Corporate Transparency Act as a constitutional exercise of congressional power. The law stands. Treasury simply decided not to enforce the part that applies to Americans, and then went further and decided to destroy the data it had already lawfully collected. Scott Bessent called it a victory for common sense and said Trump promised to cut red tape. Read that carefully. Filling out a form once, listing who owns your company, is red tape. Deleting a federal law enforcement database is not deregulation. It is destruction of evidence at scale, before anyone knew what the evidence would have shown. A rule can be rewritten by the next administration. Deleted records do not come back.

The International Operational Dialogue on Countering Cyber Scam and Forced Criminality concluded in Bangkok on Friday, launching a pioneering information-sharing platform designed by Thailand to dismantle transnational cyber scams and human trafficking networks.

AI corporations are steadily increasing malicious use risks by arguing that because the marginal risk of their model release is low, there is nothing to worry about. This “marginal risk” justification is bad for three reasons. 1. No one knows how to compute it. Which hazardous capability evaluations estimate the risks? What if the model is higher on some evals but lower on others? If it is higher on all, how much higher can each be before the marginal risk is unacceptable? Marginal risk depends on safeguards too: how do we combine capabilities benchmarks with separate refusal (and adversarial robustness) benchmarks? Do safeguardless models far from the frontier count as the baseline? How is this action-guiding when the multidimensional risk frontier shifts weekly with new releases, and when competitors' system-level safeguards are modified constantly during deployment? Evidently this was not designed to actually guide internal launch bars but to justify releases. 2. The criterion is wrong. If competitors keep one-upping each other, steadily ratcheting up risk, the marginal risk rule eventually blesses the release of fully autonomous expert-level virologists and cybercampaign orchestrators. It permits catastrophe as long as you get there gradually. 3. The law disagrees. If one of these models enables a catastrophe, the AI developer likely failed to implement reasonable safeguards against foreseeable harms. The law doesn’t let companies off the hook because someone else behaved recklessly, but the marginal risk framework does. "Everyone else is doing it" doesn't work as a safety policy. AI companies should publicly commit to fixed, empirical safety thresholds for their releases.

Cybercab is now at price and wait time parity with Uber on average. It has been raining today in Austin. Plus we got a huge geofence expansion. Still these numbers. Robotaxi is getting real. They are about to dominate Austin

Cybercab is on the floor at Auto Zürich today, the first stop of its Europe tour. Elon on what autonomy does to the car business: "now you go from like, I don't know, 20, 25% margins to 90% margins. It's insane." "For cars that are owned by customers, then we'd obviously revenue share with the customer. So the owners of the cars, I think, would earn quite a bit more than the lease cost or financing cost of the car when it's in autonomous mode." @elonmusk Auto Zürich: Nov 5–8 at Messe Zürich.
12 stops, 9 countries, through mid-January 2027. Robotaxi is already live in Fort Worth. Europe is looking at the no-wheel car in person. Interview: TOSV 𝕏 Takeover 2024 | Cybercab: @Tesla @robotaxi