Agentic AI is becoming an operating model
AI agents are moving from clever demos to workflow actors. The interesting question is no longer whether they can act, but where humans keep control, review, and accountability.
Add links in archiveWeekly AI reading notes
What is worth reading in AI this week.
A personal digest of useful AI news, sharp articles, practical tools, and security ideas worth keeping on the radar.
Current focus
Agentic AI, Shadow AI, MCP, and enterprise security
Updated
June 9, 2026
Format
Short notes with links and takeaways
This week
AI nuggets
Quick scans first, deeper reads second. Each item should explain why it matters before it sends you elsewhere.
Shadow AI is the new Shadow IT
Unsanctioned AI tools do not just expose data. They can reason over it, automate decisions, and quietly create new identity and governance gaps.
Explore the themeMCP makes AI integration reusable
Model Context Protocol is becoming a practical way to connect AI assistants to tools, context, and business systems without building every integration from scratch.
Save as a reading trackAgentic hijacking needs plain-language explanation
When assistants can browse, click, and remember, malicious pages can become instruction sources. That risk deserves practical examples, not only security jargon.
Read the security lensThe quantum cliff belongs in AI security planning
Agentic AI risk and post-quantum readiness may look separate, but both push organizations toward stronger trust, identity, and data protection discipline.
Track as a long readReading tracks
Follow the themes
AI Security
Shadow AI, agentic hijacking, governance, and Zero Trust controls.
AI Tools
Assistants, coding tools, MCP, automation, and workflow platforms.
Enterprise AI
Adoption patterns, operating models, and what matters for IT teams.
Worth Reading
Articles, reports, and posts that are useful enough to save.
Archive
Recent notes to publish
These are ready-made lanes for turning your existing AI notes into public weekly entries.
- AI security Shadow AI Is the New Shadow IT
- AI integration Model Context Protocol and reusable AI integrations
- Agent safety The rise of agentic hijacking
- Strategy The 2026 dual imperative: agentic AI and quantum readiness
Private lab
Existing project areas
The homepage now focuses on AI sharing, while the existing private dashboard areas remain reachable for your own testing.