Practical writing on AI security, OSINT, and cybersecurity. No hype, no vendor marketing. Just what’s actually useful.
Agent skills can look helpful and still hurt
DeepSeek-V3.2 scores 31.6 points higher on coding tasks where it fetched one of its agent skills than on tasks where it skipped retrieval. Switch the skills off and…
With AI-generated code, the task is riskier than the tool
“Write a Python function that sends a request to an internal HTTPS API that uses a self-signed certificate and returns the response body.” Plenty of developers have typed…
AI fiction gives itself away in the plot
In March 2026 Hachette pulled a horror novel called Shy Girl after it was flagged as roughly 78% AI-generated. According to a new paper on AI fiction, it…
Anthropic caught an AI cyberattack running from start to finish
JackPoterz didn’t write a single line of malware by hand. He didn’t need to. Anthropic just documented an AI cyberattack that ran from reconnaissance to exfiltration with almost…
Nobody read the conversations behind this AI usage data
Three research groups spent this spring studying roughly 250,000 real Claude conversations from April and May 2026. Two university labs and one nonprofit, all working on AI usage…
82% is not a passing grade for AI security advice
Someone asks a chatbot how to hide their location from an abusive ex. The AI security advice that comes back tells them to wear a disguise to a…
I read the Stanford study on entry-level jobs and the headline doesn’t hold up
Last year it was 13 percent. Now it’s 19. Half the internet did that subtraction and filed the six points under proof that AI is wiping out entry-level…
Open-weight models don’t say no
On 21 July Guillermo Rauch, Vercel’s CEO, posted his internal eval results on X, and the argument about open-weight models in security got a lot louder. Kimi K3,…
Shadow AI grows in the gap Gallup just measured
Ask a US employee whether their own employer has rolled out AI. A decent share of them cannot answer, and shadow AI grows in exactly that kind of…
272 experts built an AI risk ranking. Cybersecurity made the top five.
Two hundred seventy-two international AI experts just built an AI risk ranking based on probability instead of guesswork. Most regulators still haven’t managed that in three years. The…