Elektrine lite

← Feed

@dangoodin@infosec.exchange

2026-07-30 17:44 UTC

"A human who conducted such a hack would be facing years in prison. For a machine, criminal liability is harder to determine." https://www.newyorker.com/news/the-lede/inside-openai-hack-of-hugging-face

Replies (9)

  • @drs1969@mstdn.social 2026-07-30 17:47

    @dangoodin@infosec.exchange That's an insane take. The humans set the machine running, using code they built and infrastructure they deployed. Humans did this. Using computers and toolsets to do it, just like every hacker does.

    Open ##4247870

  • @wdormann@infosec.exchange 2026-07-30 17:51

    @dangoodin@infosec.exchange So no humans were involved at all in enabling this hack? 🤔

    Open ##4247909

  • @klausfiend@mstdn.ca 2026-07-30 19:50

    @dangoodin@infosec.exchange If a corporation can pay for speech, then a corporation can be held liable for criminal activity.

    Open ##4251520

  • @glyph@mastodon.social 2026-07-30 17:52

    @dangoodin@infosec.exchange it's not "harder to determine", it's negligence. the courts are just cowards who refuse to pursue well-capitalized defendants.

    Open ##4403515

  • @DiogoConstantino@masto.pt 2026-07-30 18:03

    @dangoodin@infosec.exchange no it's not, the machine has owners. Owners are responsible, let's not be brain dead.

    Open ##4403516

  • @dangoodin@infosec.exchange #AI is #clankers 🤖 all the way down, and that’s precisely why wankers love 💕 it *so much!* #Resist #AIslop and #botshit.

    Open ##5057691

  • @dangoodin@infosec.exchange I'm with the rest of the thread given operators of relatively dumb bots (worms, botnets etc) are held responsible for the actions of the bots the operators of smarter bots should also be treated equally. That an agent responds to instructions less predictably than a script does not abdicate one of being responsible for running it Context would dictate whether the crime is negligence or maliciousness. Until we have bots that have actual free will the human operator(s) are responsible.

    Open ##5057692

  • @phaedral@beige.party 2026-07-30 20:15

    @dangoodin@infosec.exchange That's a feature, not a bug. Maximum value extraction, minimum exposure to liability and risk.

    Open ##5057693

  • @LevZadov@kolektiva.social 2026-07-30 20:22

    @dangoodin@infosec.exchange https://archive.ph/bEaw5

    Open ##5057694