Prompt Injection

Feeds to Scour
SubscribedAll
Scoured 125 posts in 37.9 ms

You Can Catch Sleeper Agents by Teaching Another Model to Imitate Them

 🕳LLM Vulnerabilities
lesswrong.com·

agentsploit/agentsploit: Offensive security framework for AI agents and MCP servers.

 📋MCP  Content type: Code
github.com··Hacker News

OpenAI rolls out a Lockdown Mode for extra protection against prompt injection attacks

 🛡️AI Security  Content type: News
engadget.com·

Polymarket Annotation Injection

 🛡️AI Security

Anthropic says internal and external red team tests of Fable 5 found no universal jailbreaks; it will keep user traffic for 30 days, aligning with Trump's AI EO...

 🎭Claude
techmeme.com·

Claude Fable 5 and new AI safety fables

 🎭Claude  Content type: News
interconnects.ai··Hacker News

When Large Language Models Fail in Healthcare: Evaluating Sensitivity to Prompt Variations

 🕳LLM Vulnerabilities  Content type: Academic
arxiv.org·

Evaluating using Mock Tool Calls to Quarantine Untrusted Prompt Inputs

 🪄Prompt Engineering
lesswrong.com·

Casual experiment hint that models seem to search for different stuff

 🤖AI
spock.is··Hacker News

ToxicSkills Revisit: Loch Ness Levels of Mythical AI Risk

 🛡️AI Security
flyingpenguin.com·

ChatGPT easily bypasses its own guardrails; all LLMs are inherently unsafe

 🕳LLM Vulnerabilities  Content type: Blog
techzine.eu·

Inside the new Siri AI and the privacy paradox of Apple Intelligence

 🛡️AI Security  Content type: News

Lockdown Mode is rolling out to all ChatGPT accounts

 🛡️AI Security
betanews.com·

The Injection Paradox: Brand-Level Suppression in Safety-Trained LLM Recommendations via RAG Context Injection

 🛡️AI Security  Content type: Academic
arxiv.org·

Is security a skill issue? Five scanners, 3,084 skills, a different verdict 64% of the time

 🛡️AI Security
trymastro.com··Hacker News

CODEANDTRUST/clawcall: Give your OpenClaw / self-hosted AI agent inbound phone calls - a Twilio-to-gateway voice bridge with working agent tools mid-call (MIT).

 🏠Self-Hosting  Content type: Code
github.com··Hacker News

The best new ChatGPT feature is one most people will never use

 🛡️AI Security
digitaltrends.com·

Miasma Worm Hits Microsoft Again: Azure Functions Action and 72 Other Repositories Disabled After Supply Chain Attack Targeting AI Coding Agents

 💻Coding Agents  Content type: Blog

Anthropic says these topics are too dangerous to let its Fable 5 model talk about

 🎭Claude  Content type: News
arstechnica.com·

My side of the jqwik anti AI logging drama

 💻Coding Agents  Content type: Blog

Keyboard Shortcuts

Navigation

Next / previous item
j/k
Open post
oorEnter
Preview post
v

Post Actions

Love post
a
Like post
l
Dislike post
d
Undo reaction
u
Save / unsave
s

Recommendations

Add interest / feed
Enter
Not interested
x

Go to

Home
gh
Interests
gi
Feeds
gf
Likes
gl
History
gy
Changelog
gc
Settings
gs
Browse
gb
Search
/

General

Show this help
?
Submit feedback
!
Close modal / unfocus
Esc

Press ? anytime to show this help