Furong Huang's banner
Furong Huang's profile picture

Furong Huang

@furongh10,139 subscribers

Associate professor of @umdcs @umiacs @ml_umd at UMD. Researcher in #AI/#ML, AI #Alignment, #RLHF, #Trustworthy ML, #EthicalAI, AI #Democratization, AI for ALL.

Shorts

I’m so lucky to have such amazing students! 🤩 🦾🧑‍🎓

I’m so lucky to have such amazing students! 🤩 🦾🧑‍🎓

66,601 просмотров

🚨 AI companies are betting big on Web AI Agents—but they're far more vulnerable than standalone LLMs. These agents see 👀, think 🧠, and act ⚡ for you: 🕵️‍♂️ Gather personal data from emails, messages, & calendars 🧠 Plan your day, anticipate priorities ⚡ Automate tasks, execute actions, & interact with apps But what happens when they go rogue? 🔓 Our research reveals just how easy—and more importantly, WHY—Web AI Agents are dangerously vulnerable, even when built with safety-aligned LLMs. A 🧵👇

🚨 AI companies are betting big on Web AI Agents—but they're far more vulnerable than standalone LLMs. These agents see 👀, think 🧠, and act ⚡ for you: 🕵️‍♂️ Gather personal data from emails, messages, & calendars 🧠 Plan your day, anticipate priorities ⚡ Automate tasks, execute actions, & interact with apps But what happens when they go rogue? 🔓 Our research reveals just how easy—and more importantly, WHY—Web AI Agents are dangerously vulnerable, even when built with safety-aligned LLMs. A 🧵👇

14,479 просмотров

Videos

Больше нет контента для загрузки