Attacco 'overthinking' rallenta robot fino a 7 volte: vulnerabilità nei modelli vision-language episode artwork

EPISODE · Jul 15, 2026 · 10 MIN

Attacco 'overthinking' rallenta robot fino a 7 volte: vulnerabilità nei modelli vision-language

from GoYou Cybersecurity

Ricercatori del Michigan Technological University hanno scoperto una nuova vulnerabilità nei robot che utilizzano modelli vision-language. Un attacco basato su testo strategicamente posizionato può indurre l'AI a un 'overthinking' che rallenta le risposte fino a 7 volte, creando un denial-of-service. La difesa è semplice ma critica per sistemi tempo-sensibili.

Episode metadata supplied by the publisher feed · Published Jul 15, 2026

Embed this episode

NOW PLAYING

Attacco 'overthinking' rallenta robot fino a 7 volte: vulnerabilità nei modelli vision-language

0:00 10:31

No transcript for this episode yet

We transcribe on demand. Request one and we'll notify you when it's ready — usually under 10 minutes.

No similar episodes found.

No similar podcasts found.

Frequently Asked Questions

How long is this episode of GoYou Cybersecurity?

This episode is 10 minutes long.

When was this GoYou Cybersecurity episode published?

This episode was published on July 15, 2026.

Can I download this GoYou Cybersecurity episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.
URL copied to clipboard!