I clicked the link to learn what steganography mean...
Claude Code is steganographically marking requests
101–110 of 817 posts
Re: Claude Code is steganographically marking requests
#102I reported a similar system prompt injection mechanism here:
https://news.ycombinator.com/item?id=48259288
https://github.com/anthropics/claude-code/issues/62061
Looks like they just keep finding new "creative" uses for such things, as expected. I'll keep patching them out.
Re: Claude Code is steganographically marking requests
#103I used Claude Code for a month because my boss gifted me a sub and wanted me to try it. I used that month to complete a work project and then beef up my personal harness so I'd never have to deal with Anthropic (and these sorts of shenanigans) again.
Given the Anthropic shenanigans, do you trust the personal harness code it wrote for you?
It did not write it for me, I used it to add a feature I wanted. It's a pretty small and understandable codebase, in fact :)
Re: Claude Code is steganographically marking requests
#104Value judgment aside: I am a bit surprised at how sloppily they did this. I think they could've achieved the same effect while decreasing the odds of detection via reverse engineering. (This field is known as "underhanded code", coined by the Underhanded C contest: https://www.underhanded-c.org . It's a little-known "art"; little-known for probably self-explanatory reasons. There are much cleverer ways of achieving o…
Claude Code are slopmaxxxing and you're considering their "judgement"? :-)
Re: Claude Code is steganographically marking requests
#105Earlier quoted context omitted.
> This will also allow them to, instead of blocking the distillation agents, respond with a poorer result/model, i.e. this will allow them to literally commit fraud against paying customers
1st, this technique is not fraud, and fraud is a separate accusation. 2nd, paying customers can legally and legitimately be banned and monitored for breaking terms of service, which probably includes things like using the model against U.S. export restrictions.
Banning is completely different than charging for a service you're silently not providing.
Re: Claude Code is steganographically marking requests
#106What's the point of even trying to obfuscate this with such a simple method? Could at least have hidden the targeted features by storing their hashes or embedding a bloom filter or similar
In this case, this is probably not the only stereographic tattletale.
Had a competitor pull something like this with a previous employer. They were supposed to be interoperating with a standard, but they had a secret steganographic handshake, which they used to pretend that competitors products were unreliable (they had a first mover position in a smaller national market with specific requirements, so this wasn't shooting themselves in the foot). Our guys figured out the handshake and just silently implemented it. In this case, the competitor wasn't big enough to waste engineering time on multiple such hacks, but Anthropic have time (or Claude does).
Re: Claude Code is steganographically marking requests
#107The AI race right now is in a sad state. Chinese's playbook is releases open weight models and trains them on their own chips.
Anthropic pushes fear and control. But the only way to win is by innovating. China is flooding the market with cheap, good enough models, while the U.S. is building a Chinese firewall.
Re: Claude Code is steganographically marking requests
#108Non-hugged: https://archive.is/Wdhp0
Re: Claude Code is steganographically marking requests
#109Earlier quoted context omitted.
> This will also allow them to, instead of blocking the distillation agents, respond with a poorer result/model, i.e. this will allow them to literally commit fraud against paying customers
1st, this technique is not fraud, and fraud is a separate accusation. 2nd, paying customers can legally and legitimately be banned and monitored for breaking terms of service, which probably includes things like using the model against U.S. export restrictions.
So if I change my timezone to Shanghai I deserve to get banned? Or get shitty model instead of what I’m paying for?
Re: Claude Code is steganographically marking requests
#110Headline is, frankly, awful. This isn't the AI secretly doing stuff and hiding it. This is the very human Anthropic engineers trying to detect Chinese scraping via some frankly hamfisted and unimaginative URL trickery.
I didn't assume it was the AI, just that some part of the the overall Claude Code product was doing this. I didn't assume the feature was added to Claude Code without human oversight. If it was added by Claude-the-AI itself without the humans prompting it to I would still hold the humans at Anthropic responsible. Does that make you feel better?