A decade ago, the VMware Tanzu team (at that time, Pivotal) published a security manifesto, a radical call to action of sorts, via a blog post that broke the glass of traditional security culture inside of enterprise IT. The premise was faster is safer, a radical departure from the calcified, change-resistant security policies of the time.
At its core, this thesis challenged the ultimate fundamental belief of corporate security: that to protect a system, you must lock it down and stop it from changing. Traditional IT was built on the premise that change equals risk. But it also identified a greater threat: the Advanced Persistent Threat (APT). These were human adversaries who snuck into a data center and sat silently for six months, exploiting that very lack of change.
The antidote we introduced is a framework called The Three R’s of Enterprise Security: Rotate, Repave, and Repair. If you constantly shifted the ground beneath the attacker’s feet, they could never build a foundation to launch a mega-breach. This was a radical shift in culture and process, beyond just a new technical architecture. It required enablement within organizations to deliver these new outcomes. Proficiency is a byproduct of repetition; by transforming security practices from a rare, traumatic event into a routine operational cadence, organizations build muscle memory for defense. We as an industry must align with the growing consensus that vulnerability is no longer an acceptable baseline for modern software development and delivery.
Fast forward to 2026, and that philosophy is no longer just a nice-to-have architectural framework. It is a baseline requirement for survival.
The adversary has evolved. The threat landscape is no longer dictated by human hackers typing manually into terminals over weeks and months. Today, we are facing AI-enabled threat actors operating at machine speed, the only limitation being available compute. Automated vulnerability discovery, instantaneous reverse-engineering of patches, and AI agents capable of chaining minor configurations into catastrophic remote code execution (RCE) vectors have collapsed the time-to-exploit window from months to mere hours.
If your security posture relies on manual triaging and change-control boards that meet every second Tuesday, you aren’t playing defense, you’re standing still in a lightning storm.
Here is how the Three R’s have transformed for the AI era, and how VMware Tanzu Platform’s architecture delivers them natively, turning speed into your ultimate security shield.
ROTATE: Beyond static secrets to autonomous identity
The 2017 recommendation: Change database passwords, API keys, and TLS certificates every few days or hours so leaked credentials have a short shelf life.
The 2026 reality
The massive shift toward building intelligent applications powered by autonomous AI agents has turned credential management into a high-stakes battlefield. Modern AI applications don’t just query databases; they use tools, hit external APIs, make calls to scarce or expensive LLM resources, and execute autonomous loops to solve complex problems. If an AI agent exposes a hardcoded API key during an unvetted execution path, or if an attacker manipulates an agent via prompt injection, those credentials can be scraped and weaponized by adversarial AI scripts instantly.
How Tanzu Platform delivers credential management
History has proven that static credentials are an open invitation to disaster. Tanzu Platform addresses this by treating identity as a dynamic, ephemeral asset. The platform delivers out-of-the-box, zero-downtime rotation, from the infrastructure layer all the way up to the next generation of AI-driven applications:
- Automated runtime injection: Within application runtime environments like Tanzu Platform agent foundations, application credentials are dynamically injected at runtime and completely isolated from the application code and agent logs. Developers, and the AI models themselves, never see or handle raw secrets.
- OIDC as a digital passport: Tanzu Platform leverages advanced Model Context Protocol (MCP) Gateways integrated with OpenID Connect (OIDC). Every time an application or autonomous AI agent attempts to call a tool or access a data service, it must present a verifiable, short-lived digital identity. If a credential leaks, it has already expired before an adversarial scanner can register its existence.
- Platform & ingress TLS rotation: Tanzu Platform handles the heavy lifting of certificate management natively. It provisions, tracks, and rotates internal platform certificates and ingress TLS credentials before they expire—without requiring operator intervention or causing application disruption.
REPAVE: Eradicating persistence via the fleet-wide “emergency button”
The 2017 recommendation: Use automation to destroy and rebuild virtual machines and container environments on a regular, proactive schedule from a known-good state, stripping malware of its ability to persist.
The 2026 reality
Today’s polymorphic malware doesn’t need months to sit and observe; it executes lateral movement across clusters in minutes- now powered by AI. Most know that relying on perimeter security is a failed strategy. Most know they should patch their software. While many organizations prioritize security, many fail with the actionable steps of implementing these patches and managing misconfigurations at scale. The entire infrastructure stack must be treated as inherently disposable. Wiping the slate clean must happen globally, continuously, repaved declaratively from a trusted state, and—crucially—without application downtime. Your customers should never know.
How Tanzu Platform delivers painless repaving of environments from a known good state
Tanzu Platform handles this by turning platform immutability into a routine operational habit, using the proven automation of BOSH to ensure the environment adheres to a known-good state, idempotently.
- BOSH-driven rebuilds: Instead of applying incremental configuration changes to aging infrastructure, Tanzu Platform uses BOSH to idempotently recreate the underlying virtual machines hosting your applications on a rolling basis. Using a trusted, cryptographically signed release, BOSH replaces host VMs one by one, ensuring the infrastructure is always clean and free of configuration drift.
- Shattering malware persistence: Because this recreation happens on a rolling schedule, the underlying environment could be configured to be continually renewed. As application containers are gracefully evacuated and spun up fresh on these pristine VMs, any unauthorized in-memory processes, lingering malware, or hidden exploit scripts are immediately terminated and wiped from existence.
- Zero-Downtime Operations: The historical excuse for avoiding a total infrastructure repave was the fear of a massive maintenance window. Tanzu Platform eliminates this friction entirely. Because BOSH can orchestrate this as a rolling repave, the platform natively manages the routing tier and drains active connections dynamically. This allows enterprises to execute full infrastructure repaves during normal business hours with zero application downtime and zero impact to the end user.
REPAIR: Wiping Out Dependency Debt at Machine Speed
The 2017 recommendation: Apply operating system and application patches within hours of a vendor releasing them, breaking the cycle of delaying updates out of fear of breaking the app.
The 2026 reality
The moment a security patch is released publicly, malicious AI models immediately analyze the change logs to reverse-engineer an exploit, leaving organizations with a zero-day patching window. The biggest obstacle to rapid patching isn’t a lack of awareness; it is dependency debt and operational friction. In traditional environments, patching an operating system vulnerability or a middleware flaw means forcing thousands of developers to update their Dockerfiles, rebuild their images, and run them through exhaustive testing pipelines. By the time that happens, the AI threat actor has already moved in.
How Tanzu Platform delivers faster patching of apps and infrastructure without downtime
Tanzu Platform solves this by completely separating application code from the underlying operating system and runtime layers. When a CVE drops, the platform repairs the environment at machine speed using:
- Trusted OS base images & stemcells: Developers do not curate or maintain their own base operating systems. Tanzu Platform utilizes standardized, hardened OS base images (and hardened stemcells for underlying platform infrastructure). When an OS-level vulnerability occurs, platform engineers update the base image once at the platform level.
- Tanzu Buildpacks: Tanzu Buildpacks automatically handle the application’s dependencies—injecting the language runtime (such as the JRE, Node.js, or .NET), middleware, and web servers. If, for example, a critical vulnerability is found in the Java runtime, the platform team can swap out the Java buildpack centrally and push an update that automatically extends to every impacted application on Tanzu Platform. Developers don’t have to touch a single line of code or edit a single configuration file.
- The magic of “restage”: When an underlying OS or runtime vulnerability is patched, platform engineers don’t need developers to re-push their code or trigger a heavy CI/CD pipeline. Instead, Tanzu utilizes the restage mechanism. A restage tells the platform to pull the original application package from its secure internal blobstore and re-run the buildpack lifecycle. The platform automatically compiles a brand-new, secure execution droplet using the newly patched buildpack and updated OS base image, instantly rolling out predictable security fixes across the fleet with zero downtime.
- Day-zero CVE-only patches: This platform automation is fed by Broadcom’s upstream investments as the steward of Spring and RabbitMQ. Through the Tanzu Spring Enterprise Repository, enterprise customers get access to validated, CVE-only Spring patches before they hit the open source community. Combined with Tanzu Platform app portfolio assessment, which continuously scans application portfolios to help eliminate dependency debt, the platform ensures that the code you are restaging is clean, secure, and verified.
- Hardened from root to runtime: When relying on pre-built open source packages, enterprises can blindly inherit the vulnerabilities or malicious code of upstream public build infrastructure. Tanzu Platform helps eliminate this massive blind spot by securing your application building blocks and open source dependencies directly via Bitnami Secure Images. Validated to rigorous SLSA Level 3 standards, this private “built from source” pipeline completely insulates images in the environment from upstream infrastructure compromises, ensuring that the open source images you deploy to repair and refresh your application stack are verified, secure, and untampered from day one.
Systemic immunity over palliative care
Our 2017 security post noted that relying solely on security monitoring and detection tools to protect slow-moving infrastructure was like “using a table knife to eat a bowl of soup.” Nine years later, that analogy still holds up. In an era where threat actors leverage artificial intelligence to strike at machine speed, detection, alerting and selective triage is merely palliative care.
True security requires systemic immunity. By embedding rotate, repave, and repair directly into the automated fabric of Tanzu Platform, enterprises no longer have to compromise velocity to maintain security and compliance. The platform proves that even in this new era of AI-driven complexity, the fundamental law of the cloud-native ecosystem remains unchanged: faster is still safer.
