This is your Beijing Bytes: US-China Tech War Updates podcast.

Hey listeners, Ting here, your friendly neighborhood China–cyber–hack-all-the-things nerd, and Beijing Bytes is lit this week, so let’s jack straight into the mainframe.

Over the past two weeks, the US–China tech war has basically hit “hard mode.” On the cyber front, US and Canadian agencies say Chinese state-backed operators are running a long-term espionage campaign using BRICKSTORM malware to burrow into government and critical infrastructure networks across North America, living inside VMware and Windows environments like digital squatters. CISA, the NSA and the Canadian Cyber Centre describe it as multi‑year, stealthy, and very much still active. At the same time, Cisco just revealed a Chinese-linked campaign exploiting a zero‑day in Cisco Secure Email Gateway and related products; there’s no patch yet, and Cisco Talos says the attackers have been dropping persistent backdoors since at least late November. Translation: big Western enterprises are getting quietly owned.

Zooming out from individual breaches, Washington is hard‑wiring China tech controls into law. Policy analysts at Bocconi and trade publications covering chips report that the new SAFE CHIPS Act would lock in a 30‑month ban on exporting the most advanced AI processors—think Nvidia H200s and Blackwell-class silicon—to China and other “adversarial” states, stripping the Commerce Department of wiggle room on licenses. At the same time, legal analysts say the latest US outbound investment law, folded into this year’s defense legislation, tightens scrutiny on American money flowing into Chinese AI, quantum, and other “prohibited technologies,” adding reporting and potential blocking authority.

Enforcement is getting teeth too. A Commerce Department notice just slapped a 10‑year export denial order on Richard Shih for illicitly shipping US tech to restricted Chinese entities, a case compliance lawyers are calling a warning shot that China-linked supply chains will face much tougher audits. On Capitol Hill, a congressional report covered by the Associated Press and ABC News accuses China of exploiting US Department of Energy–funded research partnerships to siphon sensitive nuclear and dual‑use technologies into the People’s Liberation Army ecosystem, with investigators pushing for stricter vetting of US–China academic collaborations.

Beijing isn’t just taking punches; it’s refactoring its own codebase. JD Supra and other legal briefings note that China has finalized major amendments to its Cybersecurity Law, effective January, cranking up fines, tightening incident reporting rules, and—crucially—explicitly baking state support for AI into the law: more data, more compute, more algorithms, with “security” as the political wrapper. Separate CAC measures on cybersecurity incident reporting that kicked in last month force Chinese network operators and critical infrastructure players to classify and rapidly report attacks, giving Beijing richer telemetry for both defense and, let’s be honest, potential offensive learning.

Strategically, think of this as managed interdependence, not a clean decoupling. Scholars writing on 2025 US–China relations point out that trade and supply chains still bind the two, but AI chips, rare earths, and data flows are now chokepoints both sides are weaponizing. US export controls are nudging Chinese champions like Huawei and domestic AI chipmakers to race for self‑reliance, while Chinese cyber units probe Western infrastructure to map pressure points for any future crisis over Taiwan or beyond.

Forecast? Expect three things: more codified US restrictions that are very hard for future administrations to unwind; faster Chinese substitution in AI hardware and secure‑by‑policy infrastructure; and cyber operations that increasingly blend classic espionage with AI‑boosted tooling, hitting both governments...