LLRX September 2025 Issue – Articles and Columns

• The Trump Administration’s Continued War Against Science, Research and Public Health, Part 3 – This is a follow up to two recent articles by Sabrina I. Pacifici on the Trump administration’s relentless attacks against science, medicine and public health, government sponsored data collection and reporting, climate science, free speech, and the censorship of federally funded academic research and scholarship. The rapid fire assault against the heart of our democracy stunningly continues to escalate, per the Project 2025 roadmap operationalized under the direction of Russell Vought and Stephen Miller, fracturing our public policy, governance, the economy, muzzling the education system, and eradicating our foreign policy and diplomacy. Pacifici’s article focuses on the administration’s new actions in September 2025, documenting censorship in all sectors, across agencies, universities, corporate activities and the economy.
• Dear ChatGPT: Words Matter – Stephen Embry writes about how much we might be losing when we let AI sanitize our word choices: the difference between good and memorable often comes down to a single word. For lawyers especially, professionals whose job is communication and persuasion, completely ceding editorial judgment to algorithms that prioritize blandness over impact is a mistake. Embry states – Words matter. Don’t let robots choose yours.
• With ChatGPT, law-school instructor Sean Harrington is rebuilding student assessment for the AI era – Sean Harrington, who teaches students AI and law at the University of Oklahoma and holds both a JD and an MS in Data Analytics, saw a core problem the moment generative AI went mainstream: traditional take-home exams no longer reveal what students really know.
• The divergence of law firms from lawyers – Jordan Furlong contends right now it’s possible for an ordinary person to obtain from an LLM like ChatGPT-5 the performance of a legal task — the provision of legal analysis, the production of a legal instrument, the delivery of legal advice — that previously could only be acquired from a human lawyer. He further states he’s not saying a person should do that. The LLM’s output might be effective and reliable, or it might prove disastrously off-base. But many people are already using LLMs in this way, and in the absence of other accessible options for legal assistance, they will continue to do so.
• Political witch hunts and blacklists: Donald Trump and the new era of McCarthyism – Shannon Brincat, Frank Mols and Gail Crimmins report that a modern-day political inquisition is unfolding in “digital town squares” across the United States. The slain far-right activist Charlie Kirk has become a focal point for a coordinated campaign of silencing critics that chillingly echoes one of the darkest chapters in American history. They state t”his is far-right “cancel culture”, the likes of which the US hasn’t seen since the McCarthy era in the 1950s.”
• The Operational Protocol Method: Systematic LLM Specialization Through Collaborative Persona Engineering and Agent Coordination – This paper by Dennis Kennedy introduces a systematic methodology for transforming generic Large Language Models into specialized, persistent AI advisors and helpers through structured protocol frameworks and collaborative development processes, enabling reliable human-AI collaboration for complex decision-making across professional and personal domains.
• Google wasn’t against this privacy bill, officially. Behind the scenes, it orchestrated opposition – Reporters Khari Johnson and Yue Stella Yu investigate how Google organized business owners against California legislation to force its Chrome web browser to safeguard personal data.
• Another Brilliant Idea! the Hidden Dangers of Sycophantic AI – Jordan Furlong’s article expands analysis on the already noted risks arising from lawyers using AI. Generative AI can be incredibly, and dangerously, sycophantic. This is particularly worrisome for lawyers, because if they lose intellectual skills, what will they left to offer people? Furlong notes that the similarities between lawyer thinking and AI “thinking” should be a cause for alarm within the legal profession.
• AI In Finance and Banking, September 30, 2025 – by Sabrina I. Pacifici. Five highlights from this post: A Research Agenda for the Economics of Transformative AI; Import AI 429: Eval the world economy; Do Markets Believe in Transformative AI?; AI and Task Efficiency; and Harnessing artificial intelligence for monitoring financial markets..
• AI In Finance and Banking, September 15, 2025 – by Sabrina I. Pacifici. Five highlights from this post: Goldman Sachs bankers explore limits of AI: ‘The risk is over-reliance’.; AI Agents for Economic Research; The State of AI in Financial Services in 2025 — views from our front row seats; Managing explanations: how regulators can address AI explainability; and Artificial Writing and Automated Detection.
• Pete Recommends – Weekly highlights on cyber security issues, September 27, 2025 – by Pete Weiss. Four highlights from this week: 48% of Cybersecurity Bosses Failed to Report a Breach This Year; LinkedIn will use your data to train their AI starting Nov 3; Reuters Asked AI Bots to Scam the Elderly. They Obliged; and This is the fastest way to tell if a photo is AI-generated.
• Pete Recommends – Weekly highlights on cyber security issues, September 20, 2025 – by Pete Weiss. Five highlights from this week: NIST says that there are three main ways to sanitize data; Google misled users about their privacy and now owes them $425m, says court; USAi tool lets agencies test for AI biases, GSA official says; FBI warns of cybercriminals using fake FBI crime reporting portals; and Morgan Stanley fined $35m after hard drives sold with customer info still on them.
• Pete Recommends – Weekly highlights on cyber security issues, September 13, 2025 – by Pete Weiss. Five highlights from this week: How much is the Facebook settlement payout per person?; New Study: How Often Do AI Assistants Hallucinate Links? (16 Million URLs Studied); It Is Happening – Don’t cross a U.S. border without a “perfect burner phone”; When typing becomes tracking; and Study reveals widespread silent keystroke interception.
• Pete Recommends – Weekly highlights on cyber security issues, September 6, 2025 – by Pete Weiss. Five highlights from this week: Ice obtains access to Israeli-made spyware that can hack phones and encrypted apps; Amazon to Enter the AI Agent Race in a Big Way, Internal Documents; Selling Surveillance as Convenience; Wired, Business Insider Editors Duped By Completely Bogus ‘AI’ Using ‘Journalist’ Who Made Up Towns, People That Don’t Exist; and Verizon Finally Restores Service in Most Areas After Day-Long Outage.

LLRX.com® – the free web journal on law, technology, knowledge discovery and research for Librarians, Lawyers, Researchers, Academics, and Journalists. Founded in 1996, published monthly by a solo, independent, self funded owner, editor, publisher, researcher and writer.