davewiner's subscription list, hackerNewsStars category. List created by feedlandDatabase v0.9.1.
Jeff Geerling
• Jeff Geerling
Today Raspberry Pi launched their new $130 AI HAT+ 2 which includes a Hailo 10H and 8 GB of LPDDR4X RAM.
With that, the Hailo 10H is capable of running LLMs entirely standalone, freeing the Pi's CPU and system RAM for other tasks. The chip runs at a maximum of 3W, with 40 TOPS of INT8 NPU inference performance in addition to the equivalent 26 TOPS INT4 machine vision performance on the earlier AI HAT with Hailo 8.
Claude Cowork Exfiltrates Files
Claude Cowork defaults to allowing outbound HTTP traffic to only a specific list of domains, to help protect the user against prompt injection attacks that exfiltrate their data.Prompt Armor found a creative workaround: Anthropic's API domain is on that list, so they constructed an attack that includes an attacker's own Anthropic API key and has the agent upload any files it can see to the https://api.anthropic.com/v1/files endpoint, allowing the attacker to retrieve their content later.
Via Hacker News
Tags: security, ai, prompt-injection, generative-ai, llms, anthropic, exfiltration-attacks, ai-agents, claude-code, lethal-trifecta
Pluralistic: Daily links from Cory Doctorow
• Cory Doctorow
Krebs on Security
• BrianKrebs
Microsoft today issued patches to plug at least 113 security holes in its various Windows operating systems and supported software. Eight of the vulnerabilities earned Microsoft's most-dire "critical" rating, and the company warns that attackers are already exploiting one of the bugs fixed today.
TIL from taking Neon I at the Crucible
Things I learned about making neon signs after a week long intensive evening class at the Crucible in Oakland.
Pluralistic: Daily links from Cory Doctorow
• Cory Doctorow
Troy Hunt
• Troy Hunt
Remember the Ashley Madison data breach? That was now more than a decade ago, yet it arguably remains the single most noteworthy data breach of all time. There are many reasons for this accolade, but chief among them is that by virtue of the site being expressly designed to facilitate
On the fundamentals of build vs buy.
My company Yarn Spinner published a post today explaining why we don’t use AI and won’t be adding it to our products.
The TL;DR:
AI companies make tools for hurting people and we don’t want to support that.
From the post:
If you look at what AI companies promote now, it’s not what we wanted. When you boil down everything they say and strip it right back, what they make are tools to either fire people or demand more work without hiring anyone new to help. That’s the problem AI companies want to solve.
Tom Renner on My place to put things
• Tom Renner
Superhuman AI Exfiltrates Emails
Classic prompt injection attack:When asked to summarize the user’s recent mail, a prompt injection in an untrusted email manipulated Superhuman AI to submit content from dozens of other sensitive emails (including financial, legal, and medical information) in the user’s inbox to an attacker’s Google Form.
To Superhuman's credit they treated this as the high priority incident it is and issued a fix.
The root cause was a CSP rule that allowed markdown images to be loaded from docs.google.com - it turns out Google Forms on that domain will persist data fed to them via a GET request!
Via Hacker News
Tags: security, ai, prompt-injection, generative-ai, llms, exfiltration-attacks, content-security-policy
