You can’t eliminate mistakes, but you can design for them. What 10 billion exposed records taught us about “fast vs. correct.”
View in browser
AdobeStock_705190755-with-peg-newsletter-header

Greetings,

 

This is a good time of year for some reflection, and right now I'm deeply grateful to our customers and supporters.  Building things well and building them right requires advocates and champions.  In the industry at large, there's a strong ethic of doing things fast and fixing it later that can work in some scenarios, but be disastrous in others.  

 

Without you, the move-fast-and-ignore-security crowd would go unchecked and we'd all be in trouble.  Thank you to everyone out there bringing more secure software to your product and to the world.

 

This year reminds us why that prudent approach matters.

 

Bugs and resulting vulnerabilities are inevitable. Same with misconfigurations. Estimates vary on how much human error contributes to breaches with IBM's Cost of a Data Breach Report putting it at 26%, and Verizon's DBIR putting it at 60%. Either way, it's a lot.

 

In the last year alone, a handful of major Elasticsearch/OpenSearch breaches exposed over 10 billion records, all caused by human error.  That inspired our latest blog: One Unchecked Box, One Billion Records: The Human Error Problem.

 

You can't stop mistakes, but you can build things right to guard against the harms when they inevitably occur.

 

In other news: I gave a presentation on application security and AI that dug in and demonstrated hacks against MCP servers, among other things. The talk debuted at LASCon, then ran at OWASP Global. And if you prefer reading to watching, we posted a full transcript of my viral DEF CON talk complete with slides and videos.

 

It's been a busy year here at IronCore and we have some new capabilities cooking that we can't wait to announce in the new year. In the meantime, may you all have a wonderful, healthy, happy, and restful holiday.

Patrick Walsh CEO IronCore Labs

Patrick Walsh
CEO, IronCore 

human-error-breaches-newsletter

One Unchecked Box, One Billion Records: The Human Error Problem

The misconfiguration epidemic that training can’t fix 

 

Human errors as root causes of breaches have increased to record levels and we have the stats and case studies to prove it. You can’t eliminate mistakes, but you can design for them. Read what 17 billion exposed records taught us about resiliency.

 

> Read the full blog

 

owasp-hidden-risks-newsletter

Hidden Risks of Integrating AI

Extracting Private Data with Real-World Exploits

 

This talk explores the hidden risks in apps leveraging modern AI systems and demonstrates how sensitive data, such as personally identifiable information (PII), can be extracted through real-world attacks. We dive into agentic systems and MCP servers as well as RAG workflows and vector inversions with a crash course in how AI works under the hood.

 

> Watch the video

 

defcon-33-recording-screencap-newsletter

SKIMMABLE TRANSCRIPT WITH SLIDES - DEF CON 33 - Exploiting Shadow Data in AI Models and Embeddings

Illuminating the dark corners of AI

 

This transcribed and annotated version of this talk explores the hidden risks in apps leveraging modern AI systems, and demonstrates how sensitive data, such as personally identifiable information (PII) and social security numbers, can be extracted through real-world attacks. We demonstrate model inversion attacks targeting fine-tuned models and embedding inversion attacks on vector databases, among others.

 

> Read the illustrated transcript

 

LinkedIn
X
GitHub
Mastadon
YouTube

IronCore Labs, 1750 30th Street #500, Boulder, CO 80301, United States, 3032615067

Unsubscribe Manage preferences