A reminder to folks playing around with BabyAGI and AutoGPT, from a roboticist. Using it to retrieve information is one thing. But if you are actually coding it to DO THINGS and take action on the world, *you* are ultimately responsible for its actions.
SECURITY
-
Building Proactive Defenses Against Rogue AI Systems
By
–
I don’t want to stop the development of AI. But I think we need to build proactive defense against rogue AI, just like we build antivirus and firewall SW.
-
AI Agents Privacy Solution Through Inter-Agent Communication
By
–
Mini Yohei Says: One way to solve the privacy/data-leaking problem of someone's AI agent communicating with someone else's AI agent is to have the two agents communicate thro…
-
Securing Billions of AI Agents: Proactive Defense Strategies
By
–
Any ideas on how to increase security (proactive defense) in a world where there are billions of AI agents (AutoGPT, BabyAGI) running in apps and servers, and we don’t know what they are talking about? Some kind of antivirus against AI agents?
-
Cybercrime Costs Reach $10.5T: Automation and Analytics Solutions
By
–
“Cybercrime costs are predicted to reach $10.5T annually by 2025! What’s driving this trend? Dr. Boaz Gelbord, CSO at @Akamai
, says digital currency, jurisdictional challenges, and system vulnerabilities. Learn how automation, analytics, and architecture can help protect your -
AI’s Inhuman Advantage in Modern Warfare
By
–
#AI’s Inhuman Advantage in War
#RuleoftheRobots
@WarOnTheRocks https://
warontherocks.com/2023/04/ais-in
human-advantage/
… -
OpenAI ChatML Claims to Mitigate Injection Attacks
By
–
correct, although OpenAI does claim ChatML should help mitigate injections halfway down this page https://
github.com/openai/openai-
python/blob/main/chatml.md
… -
ChatML Format Helps Prevent Prompt Injection Attacks
By
–
yeah based on OpenAI's messaging, ChatML (which this system/user format is) is supposed to help against injections scroll halfway down this page https://
github.com/openai/openai-
python/blob/main/chatml.md
… -
System Prompt Leaks and GPT Model Vulnerabilities
By
–
this is not only applicable to GPT-4… GPT-4 is actually the best at countering these sorts of attacks if you can believe it, back before ChatML was introduced and when it was only GPT-3, it was even easier to leak the system prompt
-
System Prompt Exposure: Security and Jailbreak Risks for LLM Integration
By
–
this poses a massive problem for customers who are wanting to integrate LLMs into their products exposing the system prompt not only hurts your perceived product security reputation but also makes it easier to jailbreak your product and produce undesirable outputs