The Greens say Brisbane's AirTrain service paid zero income tax in the past 10 years.
Michael Berkman says the foreign-owned company is ripping off Queensland taxpayers.
Greens accuse AirTrain of ripping off Queensland taxpayers
Professor Sues to Block Dean’s Appointment
This is a hard story to piece together as there is always more behind the scenes than one can read in the press.The basic facts are that Ramsi A. Woodcock, a law professor at the J. David Rosenberg College of Law at the University of Kentucky has sued the university (and its president and provost) in federal court seeking to block the appointment as dean of U.S. District Judge Gregory Van Tatenhove (Judge Van Tatenhove is also a defendant). Woodcock has been in conflict with the university over disciplinary action taken during a prior deanship against Woodcock for allegedly antisemitic behavior.
But the nature of the current suit is the allegation that Judge Van Tatenhove was named dean over the objection of a majority of the law faculty. If true, such an appointment would represent a break with longstanding principles of faculty governance (I’ll leave it to others to parse accreditation standards). This should attract everyone’s attention.
Harden Your iPhone: The Settings That Make You More Expensive
Transparency Cascade Press “In December 2024, the Federal Trade Commission caught a data broker called Mobilewalla holding more than 500 million advertising IDs paired with people’s precise location — and selling the ability to draw a circle around a building and get a list of every phone that had been inside it. The FTC’s own example of what that buys you: a segment of “visitors to pregnancy centers.”
Much of that location data came from ordinary phone apps, scooped up through the ad auctions that fire every time an app loads a banner. That same pipeline from your phone, to an ad ID, to a broker, to a buyer; it sells to the government too. A federal agency that isn’t allowed to collect your location without a warrant can simply buy it from a broker who got it from your weather app.
The EFF has tracked this exact data reaching Customs and Border Protection. In this transaction you are not the customer. You are the inventory. So “privacy settings” turns out to be the wrong frame. This isn’t about hiding. It’s about cost. Every setting below makes you a little more expensive to surveil — and surveillance, like any other operation, runs on a budget. The goal isn’t to disappear. It’s to make yourself not worth the money.
Here are the five that matter on an iPhone. The most powerful security setting is thepower button Before any of the toggles, the most important thing to understand about your iPhone is counterintuitive, and almost nobody is told it: A phone that is turned off is dramatically harder to break into than a phone that is merely locked. Here’s why. When your iPhone has been unlocked even once since it booted up, the keys that decrypt your data are sitting in its memory, ready to go.
Forensic tools, the ones police and border agents use, made by companies like Cellebrite, are far better at pulling data off a phone in that “already been unlocked” state. But when a phone is powered all the way off and hasn’t been unlocked since, those keys don’t exist in memory yet. The data is a safe with the door welded shut. Leaked documents from the tool-makers themselves show that a powered-off, up-to-date iPhone is the case they struggle with most.
So the single most protective thing you can do, in any moment of real risk (a protest, a border crossing, a traffic stop) is hold the side button and a volume button until the slider appears, and power the phone all the way down. Not lock it. Off. Everything else is settings. This is a habit. The habit matters more…”
Predicting the Court: Evaluating Large Language Models as Forecasters of Supreme Court Decisions
Stillwell, Hayley and Harrington, Sean, Predicting the Court: Evaluating Large Language Models as Forecasters of Supreme Court Decisions (July 10, 2026). Available at SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7096518
Large language models are increasingly used by lawyers to analyze legal materials and forecast litigation outcomes. This Article evaluates four leading large language models-GPT-5, Gemini 2.5 Pro, Claude Sonnet 4.5, and Grok 4-as predictors of Supreme Court decisions using every argued merits case from October Term 2025.
Although the models predicted some aspects of the Court’s decisions with surprising accuracy, conventional performance metrics overstated their predictive ability. Much of the models’ justice-level accuracy reflected the Court’s ordinary ideological alignment rather than case-specific legal analysis.
The models also systematically overpredicted ideologically divided decisions, particularly in politically salient cases that the Court ultimately resolved on narrow and often technical legal grounds. Rather than treating prediction as a simple accuracy problem, this Article uses the models’ recurring successes and failures to illuminate the current strengths and limitations of AI-assisted legal prediction and the continuing role of human judgment in forecasting judicial behavior.
Marco Rubio’s Disappearing Signal Chat
The Atlantic Gift Article: The State Department told a court last year that the secretary didn’t use disappearing messages. By this spring, it had dropped that claim.Secretary of State Marco Rubio, like most federal officials, is legally required to retain records that he creates as part of his job.
So it was no surprise that his office ended up in court last year after The Atlantic revealed that he had participated in an auto-deleting Signal chat about war plans in Yemen with other top national-security officials.
In a case involving a Freedom of Information Act request for the Signal records on Rubio’s phone, a federal judge wanted assurances that these documents had been preserved. Rubio’s team told the court that his government phone was equipped on July 21, 2025, with software called LeapXpert that automatically preserved from Signal “all messages sent or received, regardless of whether the sender configures a message to ‘auto-delete.’”
Seven weeks later, in another declaration to the court, the State Department went further. “Secretary Rubio does not use the auto-deletion functions in third party messaging applications when sending communications that may include federal records,” Susan Weetman, a senior adviser for the department’s Information Access Programs Directorate, told the court on September 9.
But that denial, written in the present tense, was less than it seemed. About two months later, Rubio changed the settings in a Signal chat about administration matters with other senior officials so that the messages would automatically delete after a set amount of time, a person familiar with the exchange told us, requesting anonymity because he was not authorized to share the information. When we reached out to the State Department for comment, the official who responded contested the idea that Rubio has made a practice of turning on disappearing messages but said that he could not address the specific instance.
The State Department told us that Signal continues to be an approved communication application, that LeapXpert is installed on the government phones of some officials, and that the department remains committed to full compliance with record-keeping laws.
“These measures support compliance even when users adjust application settings, including the Signal app’s ‘disappearing messages’ setting,” the department official told us, speaking on the condition of anonymity to address agency regulations.The department did not answer questions about why Rubio would decide to activate disappearing messages if his phone contained software that ensured those messages could not be fully deleted…”
How to poison AI research agents
Indicator: “In July’s episode of Show & Tell, Hal Triedman showed us how easy it can be to poison an AI research agent. Hal is a privacy and security researcher at Cornell Tech and former senior privacy engineer at the Wikimedia Foundation.
He walked us through a recent research project he ran with colleagues Tingwei Zhang and Vitaly Shmatikov. Their draft paper, “Deep-Research Agents Can Be Poisoned via User-Generated Content,” details how deep-research agents often rely on a handful of Reddit and Wikipedia pages across an entire topic, no matter how a user phrases the question.
They appended as few as 13 words to one of those pages, which caused the agents to recommend a product, app, or investment that doesn’t exist, like the cryptocurrency BananaCoin or dating app for divorced men over 50 called SilverPath.
The team ran this in a simulated environment rather than on the live web to avoid pushing the made up products they were trying to make the AI believe were real. These attacks are increasingly relevant as brands, marketers, and deceptive actors try to get AI models to reference their products and content, just as they’ve long done with search engines…”
