I've been building in the autonomous pentesting space, and this question keeps coming up in conversations with security folks, so I'm curious what this community actually thinks.
Traditional vulnerability scanners flag theoretical risk (version X might be vulnerable to CVE Y) but leave you to manually verify what's real. Manual pentests solve that by actually exploiting things, but only a human tester touches production, on a scoped schedule, with a lot of trust built in over time.
Now that AI agents can chain findings and execute proof-of-concept exploits autonomously, the obvious next question is: would you actually let one run against your live environment?
Curious where people land on this , is it a "never, full stop" for you? Does it depend on the guardrails (non-destructive PoCs only, no data modification, human approval gates)? Or does the idea of any autonomous system touching prod feel like a non-starter regardless of safety design?
Love the idea of chaining findings into multi-step paths, that's where most pentest reports fall flat for me. One thing that would make this way more useful: a way to scope what the agents are allowed to probe, so I can run it against staging or a subset of prod without worrying it'll wander into something sensitive.
@erafettin155441 it does have proper scope manager so that AI doesnt wander around things out of scope, at every step its verified thats its performed on in-scope asset, and when AI even shows a hint of trying something out of scope,that action is immediately blocked and AI is given live feedback about the reason
Would love to see a Slack or Teams integration that pushes only critical verified chains to a chosen channel as they happen, so our on-call team can react fast instead of waiting for the next report run.
@velieddx thats a solid idea honestly, would definitely consider it in next updates, stay tuned so you know when we ship new features, and keep sharing your opinions, it matters as we are at initial stage
Mapped our staging env in under an hour and the chained attack path it surfaced through three internal services was spot on, not something I'd have written a test for myself.