I’ve been in rooms with some of the top AI minds. Sorry to break it to you, AGI will be underwhelming! The early signs are already here, hiding in the AI tools we use every day. (see image) AGI won’t arrive as a single leap, it’s emerging in slow motion. Happy to be

Do you like the Dark Mode more or Red Team version?

Built at 35,000 ft. Improved with your feedback. 👀 Newsletter + workshop links in the comments. https://t.co/bXSoMcng5v

What should I add next? 🔹 More video? 🔹 Deep-dives? 🔹 AI/Cloud/Security resources hub? Your input shapes v2.

What’s new on the site: • Sharper story on my work with enterprise tech leaders • Home for Cloud Security Podcast + AI Security Podcast • Space for my book: AI Security Engineering (Wiley, 2026) • Updated research + keynote topics • Hub for The Inference Stack • AI Workshop

Why the refresh? AI, Cloud & Security are shifting fast — and the conversations I’m having with CTOs, CISOs & tech leaders today look nothing like last year.

✈️ 35,000 ft. One flight. Questionable life choices. Instead of watching a movie on the London → New York run… I rebuilt my entire website mid-air. 😅 Dark mode included.

So, Most AI Projects don’t get to Production because most fail in Pilot stage 👀 yet we still rush to build models… before we’re ready to govern them. From all the work that I have done in our Tech Advisory with Tech Leaders from Fortune 500 companies successfully getting

Buying AI ≠ doing AI. Without data readiness, integration, and process maturity, you’re not innovating you’re experimenting with budget. 💸

6️⃣ TL;DR: Most AI pilots fail before they start. Not from bad models from bad readiness. Start with Governance. Data. Architecture. Then build. #AIReadiness #AgenticAI #AIGovernance #EnterpriseAI #AICompliance

5️⃣ In our next workshop, we’re showing this in action — real-world use case, real compliance outcomes. 🎙️ “Beyond Pilots: Building Secure Agentic AI in Regulated Enterprises” with Ashish Rajan & Fusefy 🔗 [Register here →]

4️⃣ AI Readiness = - Governance (Why) → accountability, policies, & risk controls - Data (What) → bias mitigation, provenance, & lineage - Architecture (How) → reusable, compliant AI platform This framework turns AI from a risky experiment into a defensible enterprise

3️⃣ The fix? Flip the order. Before you write a single line of code, build "AI Readiness" the foundation for secure, auditable, and scalable AI. Think of it as your measure twice, cut once phase.

2️⃣ The same failure pattern repeats over and over: 🚫 No AI-specific GRC → biased or unexplainable decisions 🚫 No data lineage → can’t prove data origin 🚫 No guardrails → PHI leaks, prompt injection, model drift 🚫 No readiness phase → “successful” pilot fails audit review

1️⃣ Most AI pilots in regulated industries don’t fail because the tech didn’t work. They fail because the foundation wasn’t ready. When AI touches patient data, financial transactions, or PII without governance or guardrails, …it’s not a pilot, it's a liability.

🧵 Breaking: Another enterprise AI pilot just failed… before it even started. Sound familiar? 💥 Here’s why and how to stop it from happening again 👇🏾

AI FOMO is real! But rushing in without readiness just rebrands tech debt as “innovation.” If you don’t have clarity on: 1️⃣ Identity (who/what can access AI) 2️⃣ Data (quality & governance) 3️⃣ 3rd Parties (vendors & models) You’re not ready for AI yet.

A few days in New York felt different this time. Reminded me how quickly the AI story is evolving. Here's what everyone’s already building with AI 👇🏾 I met Founders over coffee, engineers at meetups, SOC teams between shifts, and CISOs over dinner. What struck me wasn’t the

Vibe coding just made it into the dictionary 😎 Collins Dictionary crowned 'it Word of the Year for 2025! 🤯 I'm guessing 2026 would be Agentic - Do you Agree? 😬 #vibecoding #ai

CTOs, CISOs, and tech leaders What are you seeing? Are teams buying AI for value or visibility? Would love your honest take. 👇🏾

You can’t buy AI maturity. You have to build it through data, process, and people. Stop chasing features. Start solving problems. 🎯