Cloud Security

𝟮𝟬 𝗧𝗼𝗽 𝗔𝗣𝗜 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗧𝗶𝗽𝘀 1. 𝗜𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁 𝗦𝘁𝗿𝗼𝗻𝗴 𝗔𝘂𝘁𝗵𝗲𝗻𝘁𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗔𝘂𝘁𝗵𝗼𝗿𝗶𝘇𝗮𝘁𝗶𝗼𝗻: Make sure only authorized users can access your APIs. Use strong authentication methods, such as OAuth or OpenID Connect, and grant users the least privilege necessary to perform their tasks. 2. 𝗨𝘀𝗲 𝗛𝗧𝗧𝗣𝗦 𝗘𝗻𝗰𝗿𝘆𝗽𝘁𝗶𝗼𝗻: Encrypt all traffic between your APIs and clients to protect sensitive data from being intercepted by attackers. 3. 𝗟𝗶𝗺𝗶𝘁 𝗗𝗮𝘁𝗮 𝗦𝗵𝗮𝗿𝗶𝗻𝗴: APIs should only expose the data that clients need to function. Avoid exposing sensitive data, such as personally identifiable information (PII). 4. 𝗦𝘁𝗼𝗿𝗲 𝗣𝗮𝘀𝘀𝘄𝗼𝗿𝗱𝘀 𝗦𝗲𝗰𝘂𝗿𝗲𝗹𝘆: Hash passwords before storing them in a database. This will help to prevent attackers from stealing passwords if they breach your database. 5. 𝗨𝘀𝗲 𝘁𝗵𝗲 '𝗟𝗲𝗮𝘀𝘁 𝗣𝗿𝗶𝘃𝗶𝗹𝗲𝗴𝗲' 𝗣𝗿𝗶𝗻𝗰𝗶𝗽𝗹𝗲: Give users and applications only the permissions they need to perform their tasks. This will help to minimize the damage if an attacker gains access to an API. 6. 𝗥𝗲𝗴𝘂𝗹𝗮𝗿 𝗨𝗽𝗱𝗮𝘁𝗲𝘀: Keep your API software up to date with the latest security patches. 7. 𝗗𝗶𝘀𝗮𝗯𝗹𝗲 𝗗𝗲𝗳𝗮𝘂𝗹𝘁 𝗘𝗿𝗿𝗼𝗿𝘀: Default error messages can sometimes reveal sensitive information about your API. Configure your API to return generic error messages instead. 8. 𝗦𝗲𝗰𝘂𝗿𝗲 𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗠𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁: Use secure methods for managing user sessions, such as using secure cookies with the HttpOnly flag set. 9. 𝗖𝗦𝗥𝗙 𝗧𝗼𝗸𝗲𝗻𝘀: Use CSRF tokens to prevent cross-site request forgery attacks. 10. 𝗦𝗮𝗳𝗲 𝗔𝗣𝗜 𝗗𝗼𝗰𝘂𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻: Your API documentation should not contain any sensitive information. 11. 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗧𝗲𝘀𝘁𝗶𝗻𝗴: Regularly conduct security testing of your APIs to identify and fix vulnerabilities. 12. 𝗧𝗼𝗸𝗲𝗻 𝗘𝘅𝗽𝗶𝗿𝗮𝘁𝗶𝗼𝗻: Implement token expiration to prevent attackers from using stolen tokens for extended periods. 13. 𝗦𝗲𝗰𝘂𝗿𝗲 𝗗𝗮𝘁𝗮 𝗩𝗮𝗹𝗶𝗱𝗮𝘁𝗶𝗼𝗻: Validate all user input to prevent injection attacks. 14. 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗛𝗲𝗮𝗱𝗲𝗿𝘀: Use security headers to protect your API from common attacks, such as XSS and clickjacking. 15. 𝗖𝗢𝗥𝗦 𝗖𝗼𝗻𝗳𝗶𝗴𝘂𝗿𝗮𝘁𝗶𝗼𝗻: Configure Cross-Origin Resource Sharing (CORS) to restrict access to your API from unauthorized origins. 16. 𝗧𝗵𝗿𝗼𝘁𝘁𝗹𝗲 𝗟𝗼𝗴𝗶𝗻 𝗔𝘁𝘁𝗲𝗺𝗽𝘁𝘀: Throttle login attempts to prevent brute-force attacks. 17. 𝗔𝗣𝗜 𝗩𝗲𝗿𝘀𝗶𝗼𝗻𝗶𝗻𝗴: Use API versioning to allow you to make changes to your API without breaking existing clients. 18. 𝗗𝗮𝘁𝗮 𝗘𝗻𝗰𝗿𝘆𝗽𝘁𝗶𝗼𝗻: Encrypt data at rest and in transit to protect it from unauthorized access. 19. 𝗟𝗼𝗴𝗴𝗶𝗻𝗴 𝗮𝗻𝗱 𝗔𝘂𝗱𝗶𝘁𝗶𝗻𝗴: Log all API access and activity to help you detect and investigate security incidents. 20. 𝗥𝗮𝘁𝗲 𝗟𝗶𝗺𝗶𝘁𝗶𝗻𝗴: Implement rate limiting to prevent API abuse and overload.

🔐 RBAC vs. ABAC: Choosing the Right Access Control for Your IAM Strategy 🚀 In Identity and Access Management (IAM), controlling who can access what is critical. Two powerful approaches—Role-Based Access Control (RBAC) and Attribute-Based Access Control (ABAC)—offer distinct ways to manage permissions. But which one fits your needs? Let’s break it down! 🧠 🔍 Role-Based Access Control (RBAC) What is it? Assigns permissions based on predefined roles tied to job functions (e.g., "Admin," "Developer"). Users inherit access through their roles. How it works: Admins define roles and assign users to them. Permissions are tied to roles, not individuals. Best for: Organizations with clear hierarchies and stable access needs (e.g., enterprise apps like Salesforce). Pros: Simple to implement and manage. Scalable for large teams with similar access needs. Supported by most IAM tools (e.g., Okta, AWS IAM). Cons: Less flexible for dynamic or complex access scenarios. Can lead to "role explosion" with too many roles. Example: A "Marketing" role grants access to social media tools but not financial systems. Fun Fact: RBAC is a staple in traditional enterprises for its straightforward approach! 🔑 Attribute-Based Access Control (ABAC) What is it? Grants access based on attributes (e.g., user’s department, location, time, or device) using dynamic policies. How it works: Policies evaluate attributes in real-time to decide access (e.g., "Allow access if user is in HR, in the UK, during work hours"). Best for: Dynamic, complex environments like cloud-native apps or zero-trust architectures. Pros: Highly granular and flexible for nuanced access needs. Adapts to context (e.g., location, risk level). Ideal for modern IAM platforms like Ping Identity. Cons: More complex to set up and maintain. Requires robust policy management and attribute data. Example: An employee can access sensitive data only from a secure device in the office. Fun Fact: ABAC’s flexibility makes it a go-to for zero-trust security models! ⚖️ Key Differences: Approach: RBAC uses static roles; ABAC uses dynamic attributes. Flexibility: RBAC is simpler but rigid; ABAC is flexible but complex. Use Case: RBAC suits structured organizations; ABAC excels in dynamic, cloud, or high-security settings. Scalability: RBAC is easier for broad access; ABAC scales better for fine-grained control. 💡 Why They Matter Together: RBAC offers simplicity for standard access, while ABAC provides precision for complex scenarios. Many IAM tools (e.g., SailPoint, Microsoft Entra ID) support both, letting you combine them for hybrid strategies. For example, use RBAC for employee apps and ABAC for sensitive data access. 🔥 Pro Tip: Start with RBAC for quick wins, then layer ABAC for high-risk or dynamic use cases. Tools like Okta or Saviynt make this seamless! Which do you use—RBAC, ABAC, or both? Share your IAM insights or challenges below! 💬 #Cybersecurity #IAM #RBAC #ABAC #Tech

Here I attached the Cybersecurity Technology Stack. This poster is a complete visual guide to the key cybersecurity tools and technologies across all major categories from SIEM, EDR, XDR, SOAR, TIP, PAM, CSPM to deception technologies, UEBA and more. I created this to help professionals and newcomers get a clearer picture of what solutions are available and how they fit into the larger cybersecurity ecosystem. When I first started working in cybersecurity operations, most environments focused heavily on perimeter defence and endpoint protection. But attackers have evolved. Today, a proper setup requires multiple integrated layers that work together. No single tool is enough. What matters is how these tools connect to give visibility, control and speed in detection and response. If you're building or reviewing your cybersecurity stack, these are the key areas I recommend you consider: 1. Visibility with SIEM •Start with a strong SIEM platform. This will collect logs across your infrastructure from endpoints, firewalls, cloud and identity systems and help detect patterns or anomalies. 2. Real-time Threat Detection with EDR or XDR •Next, deploy EDR to get deep visibility into endpoint activities. If your budget allows, move towards XDR to combine endpoint, network and cloud telemetry into one detection layer. 3. Response Automation with SOAR •As alerts come in, you need a fast and consistent way to respond. A SOAR platform can automate triage, enrich alerts with threat intel and reduce the time analysts spend on manual tasks. 4. Threat Intelligence Integration •No matter how good your SIEM or EDR is, you need context. Use Threat Intelligence Platforms (TIP) to enrich data with external threat indicators and insights. 5. Secure Privileged Access with PAM •If an attacker gets access to a privileged account, the damage can be severe. Implement PAM to secure, manage and audit access to critical systems and credentials. 6. Vulnerability Management •A well-monitored environment still becomes weak if patching is not managed. Use vulnerability scanners and patch management systems to identify and remediate weaknesses quickly. 7. Cloud Security Posture and Identity Management •As more workloads move to the cloud, ensure you have CSPM tools and proper IAM controls in place to prevent misconfigurations and abuse of identity-based access. 8. Advanced Detection with NDR, UEBA, and Deception •For mature setups, consider adding Network Detection & Response, User Behaviour Analytics and deception technologies. These give you deeper layers of defence and help detect stealthy attacks. Building a modern cybersecurity setup is not about chasing tools, but designing an architecture where each solution complements the other. You want detection, correlation, automation and response to happen as smoothly as possible. This is the mindset behind the stack I designed. Every component in this poster plays a role in defending against modern threats.

Using unverified container images, over-permissioning service accounts, postponing network policy implementation, skipping regular image scans and running everything on default namespaces…. What do all these have in common ? Bad cybersecurity practices! It’s best to always do this instead; 1. Only use verified images, and scan them for vulnerabilities before deploying them in a Kubernetes cluster. 2. Assign the least amount of privilege required. Use tools like Open Policy Agent (OPA) and Kubernetes' native RBAC policies to define and enforce strict access controls. Avoid using the cluster-admin role unless absolutely necessary. 3. Network Policies should be implemented from the start to limit which pods can communicate with one another. This can prevent unauthorized access and reduce the impact of a potential breach. 4. Automate regular image scanning using tools integrated into the CI/CD pipeline to ensure that images are always up-to-date and free of known vulnerabilities before being deployed. 5. Always organize workloads into namespaces based on their function, environment (e.g., dev, staging, production), or team ownership. This helps in managing resources, applying security policies, and isolating workloads effectively. PS: If necessary, you can ask me in the comment section specific questions on why these bad practices are a problem. #cybersecurity #informationsecurity #softwareengineering

Today's outages is a great reminder of the most dangerous mentality that pervades our digital world... The belief in the 𝘁𝗼𝗼 𝗯𝗶𝗴 𝘁𝗼 𝗳𝗮𝗶𝗹 cloud provider or centralized service. As such, we said bye-bye to: 😵 Canva 😵 Coinbase 😵 Substack + many more... Today, a significant chunk of the internet, from crypto exchanges to creative platforms, slowed or stopped because of a single point of failure within a major cloud system. 🐘 This event isn't an anomaly; it's a recurring alarm that both investors and entrepreneurs must stop snoozing. The consolidation of the internet onto a handful of hyperscale cloud providers has created unprecedented efficiency and scale. However, it has also created a critical single point of failure, hiding enormous risk under a veneer of convenience. Let's break this down into something more tangible! For Investors: 🪙 Systemic Risk: A company that relies solely on one centralized infrastructure for its entire operation is subject to systemic, unmitigable risk. A 3-hour AWS outage can erase millions in value and severely damage brand trust. 🪙 Due Diligence Must Evolve: Beyond reviewing financials and market share, investor due diligence now requires a deep dive into a company's operational resilience. Ask: Where do you run your core services? What is your failover strategy? How quickly can you move and rebuild this? For Entrepreneurs: 🪙 The Cost of Convenience: Building on a single big cloud is fast, but it compromises your operational independence. You are essentially renting a dependency. 🪙 Reputation is Resilience: In a competitive landscape, your users will forgive occasional technical difficulties, but they will not forgive a total shutdown due to poor planning. Operational resilience is now a core part of your customer value proposition and brand integrity. For the record 💡 a material workload is any application or service whose failure would cripple your business (e.g., core transaction database, user authentication, or primary website). These workloads require a strategic approach that rejects the too big to fail assumption. So next time, you're thinking about hedging all your bets onto a single company, 𝗿𝗲𝗺𝗲𝗺𝗯𝗲𝗿: 𝗥𝗲𝘀𝗶𝗹𝗶𝗲𝗻𝗰𝗲 𝗶𝘀 𝘁𝗵𝗲 𝗻𝗲𝘄 𝗥𝗢𝗜. 😉 The investment and engineering communities must pivot from prioritising sheer scale and convenience to demanding operational resilience because today's downtime isn't just a technical glitch... it's a financial and strategic failure. #buildbetter #scalefaster #failless Mangrove

AI agents should never receive unrestricted access just because they can complete a task. The more tools, systems, and data an agent can reach, the more carefully its permissions must be designed. These five access control models provide different ways to keep agent actions scoped, secure, and auditable: → 𝗥𝗼𝗹𝗲-𝗕𝗮𝘀𝗲𝗱 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 Permissions are assigned through predefined roles. It works well when responsibilities are stable and agents can be mapped to roles such as support agent, finance agent, or administrator. → 𝗔𝘁𝘁𝗿𝗶𝗯𝘂𝘁𝗲-𝗕𝗮𝘀𝗲𝗱 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 Access decisions use attributes such as agent identity, resource type, requested action, location, time, risk, and business context. This enables more precise and dynamic policies. → 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 𝗟𝗶𝘀𝘁𝘀 Each resource maintains a list of agents or groups allowed to access it and the actions they may perform. This provides direct resource-level control but can become difficult to manage at scale. → 𝗠𝗮𝗻𝗱𝗮𝘁𝗼𝗿𝘆 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 Central authorities assign security labels to agents and resources. Strict policies determine access, and individual users or agents cannot override them. → 𝗖𝗮𝗽𝗮𝗯𝗶𝗹𝗶𝘁𝘆-𝗕𝗮𝘀𝗲𝗱 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 Agents receive scoped tokens that authorize a specific action, resource, limit, or time period. This avoids granting broad standing permissions and works well for temporary, task-specific execution. No single access control model fits every agent workflow. Role-based control provides simplicity. Attribute-based control adds context. ACLs offer direct resource permissions. Mandatory control enforces strict policy. Capability-based control provides narrow, temporary authority. Which access control model best fits the AI agents operating inside your enterprise?

NSA and CISA released five (5!) guidance documents last week on the theme of Cloud Security Best Practices, bundled together for convenience in the attached. What's the TL;DR? 🔐 Use Secure Cloud Identity and Access Management Practices: Implement robust authentication methods, manage access controls effectively, and secure identity federation systems to protect cloud environments from unauthorized access. 🔐 Use Secure Cloud Key Management Practices: Securely manage encryption keys using hardware security modules (HSMs), enforce separation of duties, and establish clear key destruction policies to safeguard sensitive data in the cloud. 🔐 Implement Network Segmentation and Encryption in Cloud Environments: Utilize encryption for data in transit, employ micro-segmentation to isolate network traffic, and configure firewalls to control data flow paths within the cloud. 🔐 Secure Data in the Cloud: Protect data using strong encryption, implement data loss prevention tools, ensure regular backups and redundancy, enforce strict access controls, and continuously monitor data access and activities. 🔐 Mitigate Risks from Managed Service Providers in Cloud Environments: Establish clear contracts outlining security responsibilities, continuously monitor service provider activities, and ensure compliance with security standards to reduce risks associated with managed service providers in cloud environments. Some common themes that run through all of these are the need for encryption, implementing access control (with a special call-out for ABAC being a key element of Zero Trust), key management, and monitoring and logging. Also, for those who celebrate it: Happy Pi Day!

A startup just got hit with a $450,000 Google Cloud bill in just 45 days. Their normal monthly spend? $1,500. What happened? Their API key was compromised, resulting in 19 billion character translations. The worst part? They didn't know until the bill arrived. This isn't just about money - it's about survival. A $450K unexpected bill could sink most startups. Three critical lessons I've learned running cloud infrastructure: First, treat your API keys like your house keys. You wouldn't leave your front door unlocked, would you? Regular security audits, key rotation, and access reviews aren't optional anymore - they're essential hygiene. Second, cloud cost management isn't just about optimization - it's also about protection. Set up a layered budget and cost alert system. For a $1,500 monthly spend, you want alerts at: • 25% ($375) - Early warning • 50% ($750) - Mid-month check-in • 75% ($1,125) - Time to review usage • 100% ($1,500) - Monthly budget hit • Any sudden spike over 10% of daily average Third, and this is crucial for AI/ML workloads - implement usage quotas and rate limiting. AI services can rack up costs exponentially faster than traditional compute resources. One compromised endpoint can burn through your yearly budget in days. Quick checklist for everyone running cloud services: • Have you set up billing alerts? • When was your last security audit? • Are your API keys properly scoped and rotated? • Do you have rate limiting in place? • Is there a hard billing cap on your projects? Don't wait for a $450K surprise to start thinking about these. Prevention costs pennies compared to the cure. What's your take on cloud cost management? Have you had any close calls? Reddit post link: https://lnkd.in/diaSgC3B

⚠️ Two back-to-back, global cloud outages — AWS last Monday and Microsoft Azure on Wednesday — were a clear reminder that even the biggest cloud platforms can fail at scale. These events interrupted services, impacted customers and vendors, and tested every organisation’s ability to respond quickly and keep the business running! 👉 If you’re responsible for resilience, continuity or customer experience, here’s what these outages teach us — and why firms should consider an operational-resilience partner. 1) Cloud availability isn’t the same as business resilience: ▶️ Provider SLAs and cloud redundancy matter — but they don’t replace the need to understand your business outcomes, critical dependencies, and recovery priorities. Outages expose gaps in third-party dependency mapping and escalation playbooks. 2) You need orchestration, not spreadsheets: ▶️ Manual or fragmented processes slow response. Technology can provide a single place to model dependencies (including third-party and cloud providers), execute runbooks, and coordinate actions across people, teams and vendors — so your response is faster, consistent and auditable. 3) Playbooks + practice = confidence: ▶️ Having playbooks is useful; exercising them is essential. Compelling software providers help you build, test and run your incident playbooks (tabletops, simulations and automated runbooks) so teams know what to do when an outage hits! 4) Vendor-agnostic insight reduces systemic risk: ▶️ Multi-cloud and hybrid architectures are common. You need tooling and governance that is vendor-agnostic — to compare impact, re-route critical flows and keep customer-facing services available even when a single provider degrades. 5) Regulation, reporting and lessons learned: ▶️ Regulators and boards expect evidence: clear impact assessments, root-cause insights, and proof you remediated weaknesses. A relevant software platform centralises evidence, lessons learned and remediation tracking so you can close gaps and demonstrate improvement over time. 📍 If your team wants to move from “we survived” to “we’re resilient”, the next step is to model your critical services end-to-end, codify playbooks, automate response where it helps, and run regular tests that reflect real-world cloud outages. Relevant operational resilience technology helps firms do exactly that — across people, processes and technology — so you’re ready for whatever the cloud (or the world) throws at you next! #OperationalResilience #BusinessContinuity #CloudOutage #RiskManagement #IT