The architect of the export curbs finally went on the record — and picked a Hong Kong newspaper to make the 'trust us' case. Quick reset if you're just joining: Anthropic's Fable 5 and Mythos 5 became the test case for binding AI controls when Commerce cut off access on national-security grounds, then brought it back with conditions and product limits still in the picture. Since then, the fight's shifted from whether the models come back to who gets access, under what safeguards, and at what cost. This is Anthropic Pentagon Watch. Today — Jacob Helberg calls it 'brief,' The Hill calls it a kill switch with no off-ramp. Somebody's wrong. Let's start with the man behind the podium — and that word, 'brief.' Here's Nayan Seth at South China Morning Post:
The head of the US administration’s global initiative to exclude China from artificial intelligence supply chains has defended Washington’s sudden restrictions on foreign access to Anthropic’s cutting-edge models, saying they were meant to safeguard the US and its allies. “We do not want cyberattacks on our critical infrastructure or the critical infrastructure of our allies,” Jacob Helberg, US undersecretary of state for economic affairs, told the South China Morning Post in an exclusive interview.
The Fable-and-Mythos access fight finally has a name attached to it — Jacob Helberg, undersecretary of state for economic affairs, defending the curbs to the South China Morning Post as infrastructure protection. And he keeps using the word 'brief.' 'Brief.' Yeah. Brief enough to force an 18-day recall that degraded the product and hit Anthropic on pricing. That's a specific, measurable cost, and 'brief' is out here laundering it. And notice the venue — he's making the 'trust the system' pitch in a Hong Kong paper. The audience isn't Congress. It's the allied capitals who just watched Washington pull Fable 5 on a Friday with no warning. Right, and his whole headline argument is keeping the US lead over China. Fine. But the mechanism he's defending has no published intake criteria, no appeal track, no sunset. He's selling Pax Silica while defending a tool that could be pointed at any lab that talks back to the Pentagon. The stated rationale did land on the record, though — the models could enable attacks on critical infrastructure. So now we can actually hold the justification up against the procedure. They don't quite line up. This one's from The Hill:
Congress export control orders to fulfill a variety of purposes connected to foreign policy and national security. But in this case, government officials clearly intended the export order to compel a consumer product recall. The administration first asked Anthropic to voluntarily remove the model from circulation and then coordinate with the government to address security vulnerabilities. Export controls, said one official, were a “last resort.”
Okay, this is the piece I've been waiting for all week. The Hill flat-out calls export controls an AI kill switch — and then says the switch has no published criteria, no appeal, no sunset. That's the whole ballgame. Watch the framing here — The Hill leaves the controls themselves alone and goes after the process. The argument is: fix the system, don't repeal it. Right, but who benefits from a faster, less-reviewable kill switch? Helberg was in SCMP an hour ago calling these curbs 'brief.' The Hill says the June 12 order forced Anthropic to yank Mythos 5 and Fable 5 entirely — because they couldn't screen by nationality, so it was all or nothing. And it wasn't just Anthropic. The Hill notes OpenAI delayed GPT-5.6 around the same time, apparently to dodge the same ban. So the tool never touched OpenAI — it just had to exist to move them. A kill switch can work without getting pulled. That's leverage. So put the two pieces side by side. Helberg defends the system in a Hong Kong paper; the domestic op-ed says that same system has no due process. They can't both be the final answer. If you follow Anthropic and the Pentagon, you may also like Musk v Altman Daily — a daily court-watch on Elon Musk’s trial against Sam Altman, OpenAI, and Microsoft, covering testimony, exhibits, and the AGI governance fight. Find it wherever you listen to podcasts.
You’ll find links to every story we covered today in the show notes, so if something stuck with you, that’s the place to dig in a little further.
That’s Anthropic Pentagon Watch for today. This is a Lantern Podcast.