If its a targeted attack, they'll know and probably "crack" it. If its just data trawling, you might get owned one day if they bother, but likely the env `AWS_KEY=encrypted_abc` looks reasonably close to `AWS_KEY=unencrypted_abc` and any API key probing will just look like it got rotated already.
I think siblings are correct that the more widespread this is, the less effective once every dump is run through a "also look for keys in the binary" step.
If you want to be really paranoid you can make it an online process: application server makes an HTTPS request to a key server on different infrastructure to get the decryption key, then decrypts the secret. Use an IP whitelist.
That way read access to your infra provider is not sufficient, it requires an active component.
Conclusions: Interproximal gingival inflammation may be significantly reduced by toothbrushing alone, the effect being significantly enhanced when toothbrushing is combined with dental floss.
I don't think we are on the verge of a y pestus pandemic, but you dismiss it somewhat lightly. Yes, there are antibiotics but there are also weaponized antibiotic resistant strains. Yes, rodents were key to the spread in medieval europe, but there are now weaponized pneumonic plague strains that spread airborne...
why would anyone connect with this to an unknown/untrusted machine? it's more like giving away a gun that will shoot you if you point it at your own head and pull the trigger
Because systems are made of components based on assumptions. The owners of a sandboxed LLM development environment provides isolated VM workstations, and allow access via SSH, with things like X forwarding disabled. IP connections are dynamically permitted via access tickets and time ranged 2FA.
Separately, developer becomes a VSCode user. Risk as assessed based on them developing on their Dec machine, accepted.
Developer then opens a ticket to use LLM for a project. Uses VSCode SSH Agent. Agent creates exfil tooling on workstation and reverse shells their box. Agent reads unauthorised data which possibly has LLM subversion triggers, or deletes data, maybe opens connections to C2 service. LLM agent pivots to cloud inference and continues to lateral movement.
The assumptions of systems composition are violated by VSCode. That's why its SSH Agent needs to be blocked.
this post is giving "the 1800s called, they want their smallpox victims back" energy. the entire scientific community agreed covid vaccines are safe, but I suppose you did your own research
reply