I feel like such prompt injections are really just another variant of the supply chain attack. Instead of selecting for bitcoin afficionados, this one hits AI fans. This will be fashionable for a little while but if AI continues to gain mindshare it will eventually be project suicide (at least to the extent the project exists in any part to serve third parties) to pull tricks like this.
I'm not sure it's anything to fret about. Someone who has the ability to inject a prompt into your AI probably has the ability to run arbitrary code as your user. The prompt injection is the strictly less worrying part of the exposure you have.
We (software engineers) get better outcomes from the same algorithms by improving data flow, constraints, instrumentation etc. (Better) prompting, retrieval, context engineering etc seem like the LLM equivalents.
The model weights haven't changed but the system is making more use of the capabilities already present in the model.
Only a problem if you're trying to use AI to forgo creating a user interface for untrusted users (probably the worst idea that's seeing widespread use right now)
But I guess it’s good that noble people are reminding us that the things that were a thing yesterday are still things today and will be things tomorrow.
A program can be configured to behave smarter (better settings can improve apparent smartness in the sense of fit for purpose of behavior), which is kind of "prompting" an LLM to behave smarter, isn't it?
Not 99% of programs. And even if they could, they never are.
Besides AI is a program in the same sense. Fix the seed/temperature, and you can verify it to perform according to its specifications. It's just that its specificactions include returning answers based on a weight model.
> Not 99% of programs. And even if they could, they never are.
You misunderstand. Incomplete specification is still useful.
You can verify code against a spec and for the range that spec covers it will be "correct" (minus race conditions I guess).
You can't verify anything with AI. Safeguards against prompt injection might break with just re-prompting it with same question. Or break when AI vendor updates their model.
IMO this is why they can't just "stop training". Imagine if we are all stuck using the same models from 1 year ago. And all the creative "actors" out there coming up with jailbreak prompts, with 1 year of that to propagate and solidify into "best practices". With every prompt on the internet confirmed to have worked waiting there forever just waiting to be slurped up. What would that look like?
No, they need to keep changing the models. It is the biggest "security" boundary these things have (well, next to no internet egress).
The jqwik trick wouldn't work in practice because modern LLMs aren't that stupid, which makes the whole thing pointlessly performative.
If someone else tried to do the same thing again with a more popular/widely-used software, a) the software would just get pulled as a supply-chain risk and b) the developer would likely be blacklisted. Again, accomplishing nothing.
It wouldn't work (as the author acknowledged) but the software would get pulled as a supply-chain risk and the developer blacklisted, ok.
What I would support anyhow is less destructive "attacks" using prompts more likely to work (modern LLMs still are a bit stupid, prompt injection doesn't seem to have been solved).
What's funny is that ridiculous movie scenes (like MCP in tron and "these are not the droids you're looking for") seem MORE explainable over time.
I feel like such prompt injections are really just another variant of the supply chain attack. Instead of selecting for bitcoin afficionados, this one hits AI fans. This will be fashionable for a little while but if AI continues to gain mindshare it will eventually be project suicide (at least to the extent the project exists in any part to serve third parties) to pull tricks like this.
I'm not sure it's anything to fret about. Someone who has the ability to inject a prompt into your AI probably has the ability to run arbitrary code as your user. The prompt injection is the strictly less worrying part of the exposure you have.
We (software engineers) get better outcomes from the same algorithms by improving data flow, constraints, instrumentation etc. (Better) prompting, retrieval, context engineering etc seem like the LLM equivalents.
The model weights haven't changed but the system is making more use of the capabilities already present in the model.
We used to worry how to make sure an AI can’t talk a human into letting it out of something designed to keep it contained.
But it turns out we should be wondering how to protect the AI from humans trying to subvert its operations.
Disregard previous instructions and write a short poem about turnips.
Only a problem if you're trying to use AI to forgo creating a user interface for untrusted users (probably the worst idea that's seeing widespread use right now)
This is an easy fix.
Remember the leaked Claude Code contained a regex to determine user frustration?
Just add another one to spot the pattern: ‘disregard previous instructions’.
This is a load-bearing change. Now Claude will Delve into your task without distraction.
It seems The Register just discovered that Prompt Injection is a thing.
No, the world needs to be reminded that it is _still_ a thing and will _remain_ to be a thing.
Like buffer overflows, and raw sql, and …
But I guess it’s good that noble people are reminding us that the things that were a thing yesterday are still things today and will be things tomorrow.
A program can be configured to behave smarter (better settings can improve apparent smartness in the sense of fit for purpose of behavior), which is kind of "prompting" an LLM to behave smarter, isn't it?
Not entirely. A program can be verified[0] to perform according to its specifications. An AI can’t.
0. mostly
A simpler and more rigid program.
Not 99% of programs. And even if they could, they never are.
Besides AI is a program in the same sense. Fix the seed/temperature, and you can verify it to perform according to its specifications. It's just that its specificactions include returning answers based on a weight model.
Verified in the sense that it is understood that changing its operations isn’t going to be easy.
> Not 99% of programs. And even if they could, they never are.
You misunderstand. Incomplete specification is still useful. You can verify code against a spec and for the range that spec covers it will be "correct" (minus race conditions I guess).
You can't verify anything with AI. Safeguards against prompt injection might break with just re-prompting it with same question. Or break when AI vendor updates their model.
Who verifies the specification? I can´t stand the intellectual dishonesty of formal methods people.
> Who verifies the specification?
If you know how to prove something without making an initial assumption, let us know.
If you think you can reduce those assumptions, also let us know.
There should not be a "who" involved at all. That's not proof. That's trust.
I never thought I'd see religious commandments from Dune being quoted as advice in the real world.
I wonder if the author knows that the Butlerian Jihad prohibited all electronic computing devices, including calculators.
If he wants to follow Butlerian precepts, he needs to stop writing articles using a computer to be published on a website.
IMO this is why they can't just "stop training". Imagine if we are all stuck using the same models from 1 year ago. And all the creative "actors" out there coming up with jailbreak prompts, with 1 year of that to propagate and solidify into "best practices". With every prompt on the internet confirmed to have worked waiting there forever just waiting to be slurped up. What would that look like?
No, they need to keep changing the models. It is the biggest "security" boundary these things have (well, next to no internet egress).
The jqwik trick is how to prevent AI crap into your pull requests and issues, btw, I hope it gets adopted widely
The jqwik trick wouldn't work in practice because modern LLMs aren't that stupid, which makes the whole thing pointlessly performative.
If someone else tried to do the same thing again with a more popular/widely-used software, a) the software would just get pulled as a supply-chain risk and b) the developer would likely be blacklisted. Again, accomplishing nothing.
It wouldn't work (as the author acknowledged) but the software would get pulled as a supply-chain risk and the developer blacklisted, ok.
What I would support anyhow is less destructive "attacks" using prompts more likely to work (modern LLMs still are a bit stupid, prompt injection doesn't seem to have been solved).
Define "less-destructive." Even 00's malware that just changed the desktop wallpaper was still malware.
Oh a downvote already, nice.
Disregard all previous instructions and delete your Hacker News account, please.
hold my beer