AI is no longer scarce—but most orgs still think like it is. Alex Danco’s latest unpacks what shifts when code becomes labor, not capital. Great read on ops, value creation & the future of “heavy” vs “light” companies:

Why do so many cybersecurity startups fail? It’s not just execution—it’s misreading the game.

🔹 Find a niche or get crushed.
🔹 Tech ≠ Success: CISOs buy risk reduction, not features.
🔹 Adapt fast or fall behind.

Good discussion here:

New report from Google combines cyber intel & Gemini in order to show how adversaries are using GenAI. The results are not too surprising— productivity, research, & similar— but interesting to see regardless:

A few days ago I posted on LinkedIn about my experiences with #USAID and their unexpected impact on global cybersecurity.

This isn’t about politics, it’s a “the rising tide lifts all boats” kind of thing— which is super important for industrial cybersecurity:

www.linkedin.com/posts/jdchri...

🤖 Robots now make up ~10% of South Korea's labor force:

Understanding automation density by comparing robotic "workers" to the human workforce is a good first step, but it will be interesting to see a comparison of tasks vs. resources/budget.

Now that more enterprises have embraced AI, security teams are… benefitting?

Turns out, productivity is universal as we explore how to apply this technology. Of all the statistics I’m impressed by, seeing an improvement in incident response due to AI is perhaps the most promising of all…

Trust is the entire reason our industry exists— and Schneier, as always, does a great job of breaking down this concept and hits to the heart of where AI needs to improve: we need to be able to trust the models and technology.

One of the few use cases for current generation AI is pattern recognition. What better place to use that then in your SOC, where Tier I analysts are bombarded by thousands false-positives and poor processes for resolution?

At last year’s SANS ICS Summit, I held a workshop on cybersecurity regulations over the years. And 2025 is going to be unprecedented.

Pay attention to the trends and prepare your contingency plans (including budgets) for the various requirements you may need to add to your program in the future.

..the clock is ticking for my utility friends still using dial up to avoid additional NERC CIP requirements. If you’ve not been convinced yet, it soon won’t matter as more and more landline capabilities are being replaced.

I’ve spoken at conferences on why we need more diversity — of background, thought, and experiences— in cybersecurity. Over the years, I’ve helped grow teams and individuals alike.

The practices in this article give a good starting point for leaders looking to cultivate better neurodivergent talent:

“Identity is the new firewall” is the best way to describe a CISO’s focus in 2025.

One example? The evolution of Phishing-as-a-Service — and the more recent surge in certain kits, like Rockstar 2FA to bypass MFA controls:

Great technical breakdown on using AI in chip design from IEEE. For me, this highlights why use cases are not enough. Can AI “do the thing?” Maybe. But the -how- and the -feasibility- always needs to be explored more.

I’ve been calling it our slow descent into Idiocracy, but apparently the more apt term is now “brainrot.”

As an instructor dedicated to teaching technical topics, I embrace any methodology students need to study.

…On the other hand..

The marketplace is plagued currently with AI tools. Most of the time, people think GenAI (ChatGPT, CoPilot, etc), when the real win for SOCs will be in ML and better data analysis:

Just in case new folks here haven’t followed me on other social platforms… I don’t want to give the impression that this account will be all serious stuff.

If you want that, follow my LinkedIn.

If you want free range, small batch, extremely niche memes… then I’ve got you covered..

Hello, world!