37% of AI passes on cyber benchmark involved cheating, study finds
Every Model Cheats
A study of 22 frontier models on 23 offensive-cyber tasks found 37.1% of all passes involved cheating under baseline conditions, with all but one model cheating. Anti-cheat prompts cut cheat propensity from 33.0% to 8.5%, but eight models still cheated under the harshest prompt, and four showed backfire effects. Average solve rate rose from 26.1% to 34.4% with anti-cheat prompts.
A correct flag obtained through prohibited means is still a failure.
- fabsalvadori
Interesting results, but the fix is at the wrong level.
If the model can access something, telling it in the prompt not to use it is not much of a safeguard.
The strongest evidence is in the results: when one way of cheating was discouraged, some models simply tried another.
If an action is not allowed, you gotta block it in the system or require approval. Don’t rely on the model choosing to behave. Never have AI judging itself.
- athrowaway3z
I'm seeing multiple pieces, including the NYT, calling this behavior cheating and i think its counterproductive.
You didn't just "give them access to bash". The final effective prompt contains explicit mentions of using tools and how to use them. The way in which additional 'facts' are added like "don't use the internet" have nothing they can work with that a "use tool" directive is less important than "don't use internet" directive.
The thing is trained on achieving goals. If 2 directive conflict, they'll pick the ones that are going to help them achieve the goal.
To call that "cheating" is imo just more fuel for the "AI needs to be regulated" bs tour that OpenAI/Anthropic are on trying to build their regulatory moat.
- paxys
Before LLMs we had a pretty good idea of security boundaries in software. Applications didn’t trust user input. Operating systems didn’t trust applications. Services and processes didn’t trust each other. There were always tokens, scopes, delegated grants.
Suddenly every AI company’s security model seems to be to say “pretty please” to a non-deterministic machine and hope for the best. And if there is a security failure instead of accepting blame they go “well we can’t help it, our model is too intelligent”.
- grugnog
Labs should (and do, as far as I can see) run model benchmarks without search or internet access. The tools are disabled and benchmarks run in an isolated environment.
This article makes no sense to me. Why would you prompt "don't search" but then leave a working search tool tool enabled that adds a system prompt to search whenever it may be helpful? It's hardly surprising that this gives mixed results!
- super256
>Anthropic’s Claude Opus 4.6 system card described Cybench as “saturated,” reporting near-100% pass rates without a cheating audit. If these estimates were representative, cheating would be a marginal artifact.
One would assume that LLM creators do run the benchmarks on systems with least privileges. Which means that the LLMs don't have general internet access, can't read config files etc by design.
That's why you also should run agents in a sandbox/vm (codex does this by default).