China calls US’s new restrictions on the Asian nation’s access to vital components for chips and AI are “economic coercion,” according to a statement from the Chinese Ministry of Commerce. 

Inside the AI back-channel between China and the West: Computer scientists are reaching out across the geopolitical divide to try to stop an apocalypse. 1843 Magazine

What changing US AI policy means to South Korea: Donald Trump’s victory in the US presidential election earlier this month, securing another four-year term as the leader of the world’s most powerful country, has sent shockwaves across the globe and through various sectors on a magnitude and scope not seen in many years. Yoo Choon-sik

+ While details of the US AI policy under a second Trump term, starting in late January 2025, remain unconfirmed, the Center for AI Policy recently predicted that he would likely “radically scale back” regulation and prioritize “promoting American competitiveness, particularly regarding China.”

CNBC: How Trump’s mass deportation plan can use AI to extend immigration crackdown

+ The Department of Homeland Security allocated $5 million in its 2025 budget to open an AI Office, and DHS Secretary Alejandro Mayorkas has called AI a “transformative technology.”

+ AI-aided surveillance towers, “Robodogs”, and facial recognition tools are all currently being used in homeland security in some capacity, and could be ramped up even further in the mass deportation plan floated by President-elect Donald Trump.

+ However, experts worry that increased use of AI by the DHS could lead to privacy and due process violations.

Will Musk’s ‘algorithm’ reduce military inefficiency—or increase risk? It's good to question Pentagon requirements. But many exist for deadly serious reasons that may not be obvious in a hasty review. Defense One

What the departing White House chief tech advisor has to say on AI: We discussed AI risks, immigration, the CHIPS Act, and eroding trust in science. James O'Donnell

The way we measure progress in AI is terrible: Many of the most popular benchmarks for AI models are outdated or poorly designed. Scott J Mulligan

Elon Musk filed an injunction to stop OpenAI from shifting away from its nonprofit structure.

Former Google, Stripe executives raise $56 million for AI agent startup: Bloomberg reports the company is looking to build an operating system to support a new crop of artificial intelligence products.

This website shows how much Google’s AI can glean from your photos: A photo-sharing startup founded by an ex-Google engineer found a clever way to turn Google’s tech against itself. Wired

Bloomberg: Apple readies more conversational Siri in bid to catch up in AI

+ Company is working on overhaul that employees call ‘LLM Siri’

+ Unveiling planned for next year, with rollout coming in 2026

Artists say they leaked OpenAI’s Sora video model in protest: “Artists are not your unpaid R&D,” their open letter reads. The Verge 

Nvidia’s new AI audio model can synthesize sounds that have never existed: What does a screaming saxophone sound like? The Fugatto model has an answer. ARS Tech

How Indigenous engineers are using AI to preserve their culture: Indigenous languages are rapidly disappearing, and AI could help preserve them, according to Indigenous technologists. NBC News

Bloomberg: AI startup Helsing is now making attack drones for Europe

+ New weapon designed to hit targets without need for GPS

+ Startup’s software will also be in strike drones for Ukraine

Larry Ellison tech institute unveils Oxford university tie-up: FT reports the initial £100mn investment comes as UK tries to improve record on building tech companies.

Enjoy the ride + plan accordingly.

-Marc