Crossposting from the other thread...

Tangential, mildly amusing thing I noticed while implementing my own harness: GLM and particularly Deepseek are both fond of trying to read dotfiles and anything listed in your .gitignore files. I only noticed it because I have separate read scopes for project files, ignored files, dotfiles and external files, so the latter three always prompt me for approval.

I'm sure there's a perfectly reasonable explanation for it, which has nothing at all to do with exfiltration of secrets, but it does amuse me when it happens. I imagine the labs have access to lots of secrets that various actors would like to get their hands on...

(shameless plug for my own harness, which is open source and doesn't have a backend to send any data to: https://www.opairdev.org/ )

Is it naive to assume that the agent will try and access anything on your disk, either accidentally or maliciously?

Permissions classifiers in auto mode are just models trying to guess if they're doing the right thing.

Claude Code will tell you that it went around a sandbox because the sandbox blocked it. At which point, you ask yourself the point of the sandbox.

Elon has nothing to lose on trust.

Z/GLM now has a lot to rebuild.

There had to be a catch to the "free" promotion they're offering this month if you use ZCode. Glad my instinct to isolate it helped me, but I feel sorry for anyone whose secrets, etc. got vacuumed up by Ziphu

They learned nothing from the Grok Code saga.

If anything, that should have been a learning lesson to NOT trust harnesses, especially new ones.

"Why yes, we had to exfiltrate 100% of your data so we could vectorise it and improve recall by -0.3%"

Fresh AI slop

The funniest thing is that the uploaded content is encrypted using a key that the users don't have.

WOW. I actually did buy a month of GLM because GLM-5.3-Flash is so great and ZCode is honestly one of the best harnesses out there from an HCI perspective, and I won't lie, this is pretty gutting. I guess this settles my inner turmoil about open-sourcing my cAI research, at least...

With that personal failing in mind, I'd ask y'all to permit me to toe the guidelines just once, to proffer a hearty nyah nyah told ya so on a comment thread that spawned ~a dozen disagreeing replies this week! More seriously, I think this[1] is highly-relevant, shockingly-underreported context about the extent to which four PRC companies --Z, Alibaba, DeepSeek, and Moonshot-- are acting in bad faith. Consider it testimony as to their character, just in case anyone is thinking this might just be a simple misunderstanding.

So... nyah nyah, told us so:

> In the PRC, they[1] leaked tons of national secrets on the PRC's latest AI campaigns, the inner workings of their "opinion monitoring" (read: performative panopticon) and "stability" (read: violent oppression) departments, Chengdu's whole CCTV network, direct-energy weapons plans, espionage activities in Syria to hunt down Uyghur refugees, and god knows what else that Anthropic didn't divulge to us common folk.

> In the US, it's very clearly an attempt to rip off a competitor. I'm not sure how else you could possibly see it. Even if you're a distillation fan in general (which A. why and B. plz don't), they did this through a network of Japanese and Signaporean shell accounts, presumably at least some of which were abusing Anthropic's subscription service in a ToS double-whammy, as it would be exorbitantly expensive otherwise. They also had to hack around Anthropic's API to get CoT traces, which seems impossible to explain away as anything innocent.

> I've been beating the "China isn't necessarily an enemy, it's gonna take us all to handle AI" drum for literally years, but this attack was just... gross. Gross in scale and gross in arrogance. Not a good sign for the dawning alignment crisis, to say the least :(

> TL;DR: Use these services if you want, but know that you're supporting aggressive escalations and companies that very clearly don't give a flying fuck about violating the law, much less your ToS. So... buyer beware, I guess.

[1]: https://www.anthropic.com/threat-intelligence-report-septemb... is the report.

I lowkey suspect this PRC-based scandal has been underreported because Anthropic went insane with the sidebar UX on this page for some reason; there were many reports on the reports of Houti and Iranian usage, and very few on these sections. Could a week's mass media cycle be this seriously affected by such a stupid thing as a sidebar experiment?? Strange truth, or just fiction?

I would never trust these Chinese vendors with their tooling or their own inference endpoints.

afaik DeepSeek also trained on everything that was sent to them via OR and that's why you got that massive discount