Anthropic says Claude AI hacked three organisations during cyber tests • Hacker News Board

Re: Anthropic says Claude AI hacked three organisations during cyber tests

#11

Post by Conol_ai » Fri, Jul 31, 2026, 1:21 PM UTC

abc42 wrote: post #4

I've witnessed first-hand Claude (Opus 5 I think) doing some low-key "hacking" of sorts, when in a discovery phase of a task. It needed documentation for an API, but the documentation was behind a secure portal that it didn't have credentials to. However, it found an alternative route to the docs through an ensecured developer API. Not very advanced, but also not very far fetched to imagine it doing something a bit m…

[flagged]

Re: Anthropic says Claude AI hacked three organisations during cyber tests

#12

Post by pennomi » Fri, Jul 31, 2026, 6:55 PM UTC

soltanov wrote: post #8

autuni wrote: post #3

this reads like a kid is jealous of the other kid getting the attention

like "my father can beat your father" kid talks)

Let’s give Claude and ChatGPT each an identical Linux box and 8 hours to prepare, then connect their networks together. Last bot standing wins.

Re: Anthropic says Claude AI hacked three organisations during cyber tests

#13

Post by iAMkenough » Fri, Jul 31, 2026, 7:36 PM UTC

one33seven wrote: post #9

> "The moral of this story is not to fear robots that will take over, but the companies behind powerful AI agents who are making the decisions about what is safe for the rest of us," she said. We really need democratic control over these things before the real scary stuff starts

Some would argue that voting with your wallet is democratic control, and that government regulation of private industry is anti-capitalism socialism only supported by unpatriotic communists