Anurag Sharma, CISSP

Yesterday’s AWS outage didn’t just crash websites, it crashed assumptions. Millions of people suddenly learned that ��the cloud” is really a handful of buildings on the East Coast. When one sneezes, the whole digital world catches a cold. For lawyers, business leaders, and anyone who signs vendor contracts, this is more than tech drama. Outages reveal in real time what your agreements are actually worth. Here are three truths I was reminded of yesterday: 1. SLAs aren’t safety nets. A service credit doesn’t restore your customers’ trust. 2. Redundancy isn’t magic. “Backups” don’t mean continuity unless it’s architected and contracted that way. 3. Communication is a legal obligation. Vague “reasonable efforts” language won’t cut it when your users can’t log in. Contracts aren’t just paperwork. They are the infrastructure of trust. Contract intelligence and analytics can help make your contracts more resilient and turn these lessons into a few actionable insights before the next outage hits. -------- Olga V. Mack Building trust and creating new categories at the intersection of contract intelligence, commerce, and AI. Let’s shape the future together.

112

61 Comments

I tuned into Palo Alto Networks’ Q4 2025 earnings call expecting AI headlines. What stood out instead was how much time Nikesh Arora dedicated to Identity security — and how central it’s becoming. Here are my summarized notes on Identity from the call: 👇 1. Identity is hitting an inflection point in the next 12–24 months. (You could hear the conviction in his tone.) 2. Unlike firewalls or networks, Identity is a real-time enforcement point. Speed here decides whether breaches succeed. 3. He floated the idea of an “Identity Firewall” — real-time clearing across fragmented providers. (In response to Rob Owens from Piper Sandler.) 4. With Agentic AI rising, every user, machine, and AI agent must be treated as privileged. No more narrow PAM definitions. 5. 90% of breaches still tie back to stolen or mismanaged credentials. That stat wasn’t delivered lightly. 6. Identity today is too fragmented...100+ vendors. Consolidation isn’t a preference anymore, it’s survival, and critical to reducing breaches. 7. His framing was clear: security-first identity vendors, not SSO-only players, will own the future. (I couldn’t help but chuckle at this. 😅) 8. Agents (human + AI) both require privileged access controls. A shift few have fully absorbed. 9. Growth in PAM data and sensors will further solidify category leaders like CyberArk. The subtext was undeniable: Identity is no longer a supporting actor. It’s becoming the frontline pillar of AI + security in the enterprise. #IAM #IGA #IdentitySecurity

133

37 Comments

Sharing my AI for Security resource collection — brand new & updated. Last year I wanted to put together and share a resource for folks who, like me, don’t come from a background in data science, machine learning, or software engineering — but do come from the cybersecurity world and want to start building with LLMs. This is a culmination of the best papers, blogs, videos, and resources I’ve come across over the past two years of building with LLMs and AI agents. It covers everything from context engineering to evals, AI agents, prompt engineering, and tons of real-world examples. If you spend even a few hours going through some of these, you’ll be way ahead of most people. There’s an overwhelming amount of information out there around LLMs and AI agents — and best practices are changing every single day. This collection should give you an incredible foundation (though I still keep the really good stuff close to my chest 😉). Link to the resource is in the comments! There’s a lot in here, but if you want the highest ROI, here are my top 5: 🧠 Building AI Products In The Probabilistic Era by Gian Segato https://lnkd.in/eGpP26NY ⚙️ A Field Guide to Rapidly Improving AI Products by Hamel Husain https://lnkd.in/e-DN-m5R 🔺 The LLM Triangle Principles to Architect Reliable AI Apps by Almog Baku (vi/vim) https://lnkd.in/evb2wxZ6 🧩 Advanced Context Engineering for Coding Agents by Dexter Horthy https://lnkd.in/erzuJk3i 🎥 Context Engineering: Lessons Learned from Scaling CoCounsel by Jake Heller https://lnkd.in/eWWPd-xN Happy building!

272

18 Comments

Many SCA tools and eBPF-based tracing solutions (that don’t offer built-in code analysis) often generate a high volume of false positives and unnecessary noise. To truly prioritize what matters for your specific product and environment, code-driven detections are key. They offer the precision in scoping the impact correctly and the context needed to make the right security decisions. Example Binarly Transparency Platform output from CVE-2024-5535 (CWE-200) detection👇

26

2 Comments

𝗲𝗕𝗣𝗙 𝗶𝘀 𝗯𝗲𝗰𝗼𝗺𝗶𝗻𝗴 𝘁𝗵𝗲 𝗻𝗲𝘄 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 𝗽𝗹𝗮𝗻𝗲 𝗳𝗼𝗿 𝗺𝗼𝗱𝗲𝗿𝗻 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 Daniel Borkmann OSSNA 2025 keynote made that point clearly. eBPF has moved far beyond packet filtering into a general-purpose way to extend the Linux kernel safely, without changing kernel code. Meta runs more than 100 eBPF programs in its production load balancer. Android uses it across billions of devices for security enforcement and traffic control. With Netkit, Meta reduced namespace resolution latency from 12 seconds to 100 milliseconds. The same model is now being extended to virtual machines. Through KubeVirt and Netkit, eBPF enables a shared layer of observability, policy, and performance across containers and VMs. Borkmann also highlighted how companies like Meta, Google, and Valve are using eBPF-based schedulers in production to optimize workloads without kernel patches. AI profiling and TCP pacing are emerging areas. With verified execution, cross-platform compiler support, and growing real-world adoption, eBPF is no longer experimental. It is infrastructure. Link to the full talk: https://lnkd.in/ePhazvfS

74

1 Comment

Enterprises lose up to $100B+ annually due to lack of security and poor segregation of permissions across the 100s of people they employ. 🤯 . . At this scale, maintaining control and security becomes critical. . . That’s why we’re thrilled to launch Role-Based Access Control (RBAC) on xPay. 🚀 This feature follows high demand from our customers - especially Late Stage Startups and Enterprises. With RBAC, you can: 🔑 Assign roles like Admin, Finance Manager, and Sales/Customer Support 📜 Customise permissions or each role ensuring sensitive actions are restricted 🗣️ Avoid passing around credentials The result? 💼 Streamlined access to your payments without compromising security. Whether you’re already scaling globally or planning to expand, xPay ensures your payment operations are secure, seamless, and enterprise-ready. Ready to take control of your global payments? Book some time here: https://lnkd.in/gf5tjFXS

82

4 Comments

The Model Context Protocol is everywhere right now, so I thought it would be interesting to look at the security posture of the most popular MCP servers. I scanned the top 161 MCP containers from Docker Hub using our open source tools (Syft for SBOM generation, Grype for vulnerability scanning). Results: 9,000 vulnerabilities including 263 criticals. That sounds dramatic, but the story is more nuanced: Most vulnerabilities are in Debian packages rated "negligible" by Debian's security team - these are deprioritized for good reason. The real issue is that many containers were built months ago and never updated. When I reached out to ClickHouse about their container having the most findings, they updated it quickly. The community response model works when we make information visible. Some observations from the data: - 36,670 NPM packages across all images (NPM dependency sprawl is real) - 97 Alpine, 61 Debian, only 2 other distros in the top containers - One container had a 600MB binary (turned out to be AI training data, not a mistake) This isn't meant to be a deep security analysis of MCP. It's a starting point for conversations. If you think I missed something or want to analyze this differently, please dig deeper - I'd love to see more research here. The fundamentals haven't changed: update regularly, minimize what you ship, know your inventory. Software ages like milk, not wine. Methodology and all the graphs in the full post: https://lnkd.in/gvQBi54t If you're working on MCP security or container hygiene, I'd love to hear your thoughts #MCP #SBOM #OpenSource #SoftwareSupplyChain

67

16 Comments

🎙️ Bridging AI and Security – My Conversation with Prabh Nair I had an amazing time joining Prabh on his show “Coffee with Prabh” ☕️ — where we dove deep into the evolving world of AI Security Architecture. We talked about: How traditional IT security architectures are breaking under the weight of AI systems. Why model security and data governance are now the new “crown jewels.” The risks of prompt injection, data poisoning, and how organizations can build attack-resilient AI pipelines. The approach to building layered, unified frameworks that bridge business strategy with engineering-driven security. Why cleaning, validating, and version-controlling data pipelines is as critical as CI/CD checks in traditional app security. Understanding the problem deeply is more important than rushing to apply solutions. AI security is not about reinventing the wheel; it’s about engineering with context, controls, and consciousness. Big thanks to Prabh Nair for the engaging discussion and for his continued efforts in spreading cybersecurity awareness across India’s community of practitioners. 🙌 🎧 Watch / Listen here: https://lnkd.in/gYSmTsUk The document i showcased : https://lnkd.in/gsR8MVqR #AIsecurity #Cybersecurity #LLMsecurity #AIdrivenSecurity #Podcast #SecurityArchitecture #Infosec

94

4 Comments