Reflecting on a Breakthrough Year

TL;DR

• Fortune 500 customer adoption confirms runtime security has been elevated to the definitive source of truth for protecting against modern attacks.
• Oligo platform expanded with Cloud Application Detection and Response (CADR), Runtime AI Security (AI-SPM and AI-DR), advanced new detections, unmatched vulnerability intelligence, and more.
• Our research uncovered industry-shaping modern threats that materially made the ecosystem safer, from AirBorne and ShadowRay 2.0 to critical vulnerabilities in AI tooling and cloud infrastructure.
• Industry validation accelerated, earning consistent recognition from analysts, CISOs, practitioners, and trusted security voices as a leader shaping the future of runtime protection.

Recapping A Banner Year

Every year in a company’s life tells a story. But some years don’t just reflect progress.

They mark the moment when a long-building shift becomes undeniable. When your core vision starts to align perfectly with the reality you thought would come. When what once sounded ambitious starts becoming the source of truth customers rely on. 

For Oligo, 2025 was the year that major signs of that change became undeniable.

From the beginning, we believed runtime security would eventually become foundational – that securing modern software would require understanding applications the way developers do: through real code execution, in real environments, with real context. 

For years, that belief ran ahead of the market… and this year, the market began to catch up.

Across customers, research, product adoption, and industry recognition, we saw runtime security move decisively from theoretical necessity to operational requirement. What had long been a conviction inside Oligo began to show up externally as demand, validation, and advocacy.

This year marked the point where our vision of runtime security stopped being just a point of view, and started becoming market reality.

Redefining What Runtime Visibility Means

One moment at the RSA Conference captured this shift clearly. While giving an elevator pitch to a CISO I had just met, our conversation was interrupted by an Oligo customer who told him they used to hate their job, but now love it because of the clarity Oligo provides.

That kind of advocacy doesn’t happen overnight. It’s the result of rethinking how runtime signals are collected, interpreted, and turned into action, all while consistently delivering outcomes that matter.

Product Breakthroughs

We are deeply technical by design, and relentlessly focused on solving real customer problems. While this year included dozens of enhancements, a few core breakthroughs defined our momentum.

Full-Stack Runtime Security

We introduced Cloud Application Detection and Response (CADR), closing the gap between application-layer exploits and workload-level threats. Powered by Deep Application Inspection, CADR delivers full attack context (from initial intrusion through post-exploitation) marking a shift from reactive visibility to proactive, full-stack runtime protection.

Runtime AI Security

We expanded our platform with Runtime AI Security, delivering real-time visibility, governance, and protection for AI models and agents in production. With AI-SPM and AI-DR, Oligo secures AI where it operates – monitoring behavior, detecting misuse, and stopping threats like prompt injection and agent abuse as they happen.

Vulnerable Function Enrichment

We delivered a major step forward in vulnerability intelligence through our ability to identify vulnerable code at the function level, uncovering 1100% more vulnerable functions than any existing CVE advisory. By enriching CVEs with LLMs and validating real-world execution in live environments, Oligo delivers findings that are precise, provably exploitable, and immediately actionable, dramatically reducing noise and MTTR.

Malicious Packages Detection

Oligo extended supply-chain security beyond CI by bringing malicious package detection to runtime. By combining continuous threat intelligence with real-time, per-package behavioral profiling, CADR detects known and zero-day malicious packages the moment they execute in production, closing a critical blind spot in OSS defense.

Observability Breakthroughs

Our runtime sensor continued to set the bar for efficiency and depth, proving that not all eBPF approaches are created equal. Visibility alone is not enough. Security requires accuracy, determinism, and meaningful context. Our work challenged shallow assumptions about observability and pushed the industry toward deeper, context-aware runtime insight.

Oligo Model Context Protocol (MCP)

Oligo MCP connected runtime-exploitable risks directly back to the source code that introduced them. By integrating with IDEs and AI assistants, MCP enables teams to trace risks, generate fixes, and open pull requests in a single workflow, shifting vulnerability management from alert triage to fast, end-to-end remediation.

Partnership with Endor Labs

Our partnership with Endor Labs unified development-time and runtime reachability in a single workflow. By correlating static call-graph analysis with live function-level execution data, teams can finally prioritize the CVEs that actually run in production.

Threat Research that Shook the Industry

Throughout 2025, our research team continuously raised the bar in identifying vulnerabilities and attack campaigns impacting the infrastructure that our customers rely on every day. 

Some of the findings that made an impact far beyond headlines are summarized below. 

AirBorne

We disclosed AirBorne, a set of vulnerabilities in Apple’s AirPlay protocol and third-party SDKs enabling severe attack chains, including wormable zero-click RCE. With 23 issues reported and 17 CVEs issued, the research highlighted how a single compromised device could become a propagation launchpad to infect Mac devices, third-party devices, and even CarPlay-enabled automobiles. 

Application Attack Matrix

We helped launchthe Application Attack Matrix: the first community-driven framework designed specifically to map real-world attacks against modern cloud applications. Inspired by MITRE ATT&CK® but focused on application-layer threats, it provides defenders a practical way to understand and counter attacks where traditional models fall short.

Vulnerabilities in AI Infrastructure

Oligo disclosed critical RCE vulnerabilities stemming from unsafe ZeroMQ and pickle usage across major AI inference frameworks, including Meta Llama Stack, NVIDIA TensorRT-LLM, and PyTorch ecosystems. The research exposed systemic risk caused by code reuse, and reinforced the need for runtime protection inside AI frameworks themselves.

We also disclosed a critical RCE in Anthropic’s MCP Inspector (CVE-2025-49596), showing how browser-based attacks could lead to full developer machine compromise. Anthropic addressed the issue by enabling secure defaults, reducing exposure across the MCP ecosystem.

ShadowRay 2.0

In 2025, Oligo uncovered ShadowRay 2.0, an active global campaign exploiting the disputed Ray flaw (CVE-2023-48022) to hijack exposed AI clusters into a self-propagating botnet. Attackers used DevOps-style infrastructure and AI-generated payloads to adapt and scale, showcasing how AI can be turned against itself in a campaign that spreads autonomously. 

Fluent Bit Vulnerabilities

We disclosed a critical chain of five Fluent Bit vulnerabilities enabling cloud compromise, from authentication bypass to RCE. Because Fluent Bit is foundational cloud infrastructure, the findings showed how weaknesses in “plumbing” components can lead directly to infrastructure takeover and attacker stealth. We worked with AWS on coordinated disclosure and published concrete mitigations.

Momentum Recognized Across the Industry

As our platform matured and research accelerated, the industry took notice. 

To recap some of the moments of recognition from this year that we are most proud of:

• Named an AI Security Innovator, Leader in CADR, and Innovator in Cloud Security by practitioner-focused research firm Latio.
• Showcased as a Leader in Application Detection and Response (ADR) by Frost & Sullivan.
• Selected as a SINET16 Innovator by a group of over 100 CISOs and risk executives.
• Named to the Fortune Cyber 60 for the second year in a row. 

These milestones reflect trust, traction, and alignment with where the industry is heading.

Building the Foundation for the Next Iconic Cybersecurity Company

What stands out most about 2025 isn’t simply the volume of what we shipped. It’s what it signaled.

Every product advancement deepened runtime understanding. Every research disclosure raised the bar for the industry. Every customer win and analyst validation reinforced the same message: runtime security is no longer optional.

Runtime security is inevitable. 

As environments become more distributed, dynamic, and AI-driven, security must operate in real time, with full execution context. That is the reality modern organizations now face, and the reality Oligo has been building for from the start.

2025 was the year the next evolution of runtime security broke through.

In 2026, we won’t just participate in that evolution. We will define what leadership looks like in it.

To our employees, customers, partners, and community: thank you for helping turn conviction into momentum, and momentum into meaningful change.

While it’s easy to be proud of the progress, we are far from done.

Back to work.

‍