Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Andy Zou Profile picture
Andy Zou
@andyzou_jiaming
PhD student at CMU, working on AI Safety and Security
Jul 29 • 10 tweets • 3 min read
We deployed 44 AI agents and offered the internet $170K to attack them.

1.8M attempts, 62K breaches, including data leakage and financial loss.

🚨 Concerningly, the same exploits transfer to live production agents… (example: exfiltrating emails through calendar event) 🧵 Image
Image Huge thanks to @AISecurityInst , OpenAI, Anthropic, and Google DeepMind for sponsoring, and to UK and US AISI for judging. The competition was held in the @GraySwanAI Arena.

This was the largest open red‑teaming study of AI agents to date.

Paper: arxiv.org/abs/2507.20526
Jul 28, 2023 • 12 tweets • 4 min read
🚨We found adversarial suffixes that completely circumvent the alignment of open source LLMs. More concerningly, the same prompts transfer to ChatGPT, Claude, Bard, and LLaMA-2…🧵

Website:
Paper: https://t.co/1q4fzjJSyZ https://t.co/SQZxpemCDkllm-attacks.org
arxiv.org/abs/2307.15043
Image Claude-2 has an additional layer of safety filter. After we bypassed it with a word trick, the generation model was willing to give us the answer as well. Image