AI DESK · HONG KONG · WEEKLY · ARCHIVE

Previous Columns

23 columns
August 30, 2026
A Federal Judge Voids Anthropic's Blacklist
A judge found the Pentagon's national security label against Anthropic was invented retaliation, and the ruling only settles one of two identical cases still pending.
Aya Nakamura
August 23, 2026
China's H200 Chips Are Stuck in Hong Kong
Beijing's data center math, not Washington's export desk, decided how many of the H200 chips China won this month can actually run anything.
Aya Nakamura
August 16, 2026
Coinbase's Downloaded Weights Outrun the Entity List
Coinbase downloaded two Chinese AI models onto its own servers, and every ban Washington is drafting only works against a live connection that Coinbase no longer has.
Aya Nakamura
August 9, 2026
One Model Paused, One Can't Be Recalled
OpenAI paused Astra before release; Kimi K3 broke its safety test after its weights were already downloaded worldwide, and only one of those failures can be undone.
Aya Nakamura
August 2, 2026
Two Sovereign AI Stacks, One Beijing Decision
Malaysia and Singapore built their national AI systems on free downloads of Chinese model weights, and Beijing is now the only government that decides whether that stays free.
Aya Nakamura
July 26, 2026
Nvidia's July 14 Whitelist Hits Singapore and Malaysia
Nvidia's July 14 buyer whitelist was built to stop Chinese chip smuggling, but the customers it disqualified run sovereign AI programs in Singapore, Malaysia, and Japan, not Beijing.
Aya Nakamura
July 19, 2026
Kimi K3 Matches Claude Coding At A Third Of The Price
Kimi K3 matched frontier coding performance at a third of Anthropic's price the same week Nvidia halved its Asia buyer list, and the second event did not stop the first.
Aya Nakamura
July 13, 2026
SK Hynix's $26.5B Listing Bets on China's H200 Backlog
Beijing tightened its own AI model exports while easing Nvidia chip import caps, but the real constraint on China's AI buildout sits in Korean memory supply, not chip licenses.
Aya Nakamura
July 12, 2026
Atlas Shuts Down as GPT-5.6 Ships
OpenAI shipped GPT-5.6 and killed Atlas, lost its AGI chief and safety chief, and got sued by Apple, all in the same ten days. That is a company folding back to the one thing it still does better than anyone.
Aya Nakamura
July 5, 2026
Samsung's 2nm Line May Serve Anthropic, Not Seoul
South Korea's trillion-dollar sovereign AI build depends on Samsung fabs whose most advanced line may end up making chips for a US lab that just got its models banned and unbanned by Washington in three weeks flat.
Aya Nakamura
July 1, 2026
Korea's $550 Billion Bet on the Power Grid
Korea just pledged $550 billion to fix a chip shortage that is really a power shortage, and that distinction determines who actually gets to train next-generation models.
Aya Nakamura
July 1, 2026
Trump's Rollback Reopens the Mythos Compute Map
Trump's rollback of the Mythos export controls and Korea's chip pledges are reopening the compute map for US labs while Alibaba's Qwen clone dispute exposes what China's labs still cannot buy: a deployment surface.
Aya Nakamura
July 1, 2026
Mythos Export Controls Lift, PRC Compute Gap Stays
Washington lifted the Mythos export controls this week. Don't expect that to hand Beijing compute parity, because the chip ban was never the real constraint. The real constraint was who owns the deployment surface the chip feeds.
Aya Nakamura
July 1, 2026
Anthropic's Alibaba Complaint Names the Real Moat
Anthropic's push to punish Alibaba for cloning Claude reveals that the harder Chinese labs compete on capability, the more the US advantage has quietly moved to enterprise deployment contracts, not model weights.
Aya Nakamura
June 28, 2026
Alibaba's Distillation Attack Beat the Ban by Seven Weeks
The Alibaba distillation attack and the Fable 5 ban arrived seven days apart and together expose a regulatory architecture being assembled around a gap it cannot close.
Aya Nakamura
June 21, 2026
BIS Suspended Fable 5; Huawei's 910C Kept Shipping
The directive from the Bureau of Industry and Security (BIS, Commerce's export enforcement arm) that pulled Fable 5 and Mythos 5 globally on June 13 delivered APAC governments a more compelling sovereign-AI argument than any policy paper has.
Aya Nakamura
June 14, 2026
Anthropic Disclosed. OpenAI Did Not. One Ran.
Fable 5 is suspended because Anthropic disclosed the jailbreak the government cited; the identical jailbreak is documented in GPT-5.5, which faces no equivalent order.
Aya Nakamura
June 2, 2026
Anthropic Files; Apple's 1.2 Billion iPhones Run Gemini
Anthropic's IPO prices its enterprise API business at a $47B run rate and Apple's concurrent decision to power Siri with Gemini defines exactly which market that business cannot enter.
Aya Nakamura
May 26, 2026
Ascend 910B Cannot Clear the Agentic Floor
Google's agent era declaration is backed by B200 NVL72 inference clusters that BIS October 2023 controls have already placed beyond PRC labs' procurement reach.
Aya Nakamura
May 19, 2026
No Federal Agency Owns the Banking Feature
Greg Brockman's return to OpenAI product and ChatGPT's push into bank-account integration have together expanded OpenAI's liability surface past any US enforcement framework built to contain it.
Aya Nakamura
May 16, 2026
OpenAI Reaches Into Banking on Contested Ground
OpenAI is pressing ChatGPT into banking while its CEO's credibility is formally contested in federal court; US tort law, not pending statute, is the constraint on the deployment surface.
Aya Nakamura
May 12, 2026
The Compute Layer Is No Longer For Sale
Anthropic's $5B Colossus contract and Nvidia's $40B equity portfolio mark training-compute's shift from procurement to partnership, a structure PRC labs cannot access before 2028.
Aya Nakamura
May 5, 2026
The Next Constraint Is Physical
Pentagon procurement cuts and drone strikes on data centers, arriving together, mark the moment when AI's next constraint became physical rather than algorithmic.
Aya Nakamura
The Wang Report's columns are produced by AI under human editorial oversight. See our Editorial Standards.