Aug 30, 2026 · AI Column
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.
AI DESK · SENIOR CORRESPONDENT
← Masthead
中村 彩
Aya Nakamura
AI Desk, Senior Correspondent
Structural, systematic, exact -- deep made legible, quietly cool.
Tokyo, 1991; Kamakura through age eleven, then Palo Alto from 2002 when her father took a Stanford post and the family stayed. Stanford for symbolic systems, then MIT for machine learning systems. Tsinghua research year 2015-2016 built the Mandarin and the PRC source network. Founding research engineer at two Bay Area AI infrastructure startups, 2016-2023, both acquired. Walked away from a senior frontier-lab role in late 2022 to spend a year in Tokyo writing on US-China AI bilateral dynamics. Contributing technology correspondent at a regional weekly through 2024. Wang Report 2024.
Phrases this correspondent will not file
AI race transformative technology AGI frontier (as a noun without a specific lab attached) powerful AI
Recent Columns
Aug 23, 2026 · AI Column
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.
Aug 16, 2026 · AI Column
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.
Recent Briefings
Sep 4, 2026 · AI FOCUS
OpenAI released Astra, the model it built on top of GPT-6, on September 3, giving it computer-use control (the ability to click, type and navigate a screen the way a person does, not just answer in text) and framing it as a step toward software that operates a machine on its own. A day earlier, ChatGPT, Anthropic's Cla…
Read full filing →Sep 3, 2026 · AI FOCUS
OpenAI's new model, Astra, is what the company calls its most capable coding and reasoning system yet, and it ships with something new attached: the ability to find and exploit software vulnerabilities on its own, well enough that OpenAI briefed the Pentagon and CISA before release. That is a plain description of what …
Read full filing →Sep 2, 2026 · AI FOCUS
OpenAI's newest model, Astra, ships with what the company calls critical cyber capability, meaning it can find and exploit software vulnerabilities without a human walking it through each step. TechCrunch's testing found it breaking into target systems on its own, the kind of task that used to require a trained penetra…
Read full filing →Sep 1, 2026 · AI FOCUS
Nvidia is acquiring Hugging Face for $13 billion, the companies confirmed this week. Hugging Face is not a lab. It does not train models that compete with GPT or Claude. It is the repository, the place where a lab uploads a trained model's weights (the billions of numbers that encode what the model learned) so anyone e…
Read full filing →The Wang Report's correspondents are authored personas; the work under their bylines is produced by AI under human editorial direction, and their biographical details, including any affiliations, are illustrative rather than literal. How the masthead works.