Filigran, the European open-source threat management company, today announced Attack Chaining, an attack-path validation capability that automates penetration testing and red teaming with OpenAEV, the company’s adversarial exposure validation product. The capability arrives today as part of OpenAEV v3.

This press release features multimedia. View the full release here: https://www.businesswire.com/news/home/20260901146373/en/

Adversaries do not run techniques in isolation. They chain them: reconnaissance surfaces a target, a credential dump yields a password, that password opens the next machine, and each step depends on what the last one uncovered. This is no longer a hypothesis; these chained vulnerabilities are already being reported. Atomic tests and predefined scenarios identify specific weaknesses well, but neither adapts to what an attacker discovers along the way.

Filigran’s State of Threat Management report, an independent global study of 550 security decision-makers and practitioners, found that 97% of organizations have difficulty determining whether their exposures are actually exploitable, and 88% still rely on manual processes for offensive attack simulation.

“Security validation has to evolve with the way attackers operate. The goal is no longer just to prove that we can block individual techniques; it is to understand whether those techniques can be combined into a path that leads to a real compromise,” said Julien Richard, co-founder of Filigran. “As adversaries become more adaptive and increasingly use AI to move faster, we need validation that can keep pace. Attack Chaining brings that level of realism and adaptability into continuous exposure validation, giving security teams a much clearer picture of the resilience they actually have.”

Attack Chaining discovers any finding—a credential, an open port, a file share, a set of permissions—and feeds it as the input to the next action, continuing, branching, or stopping the moment a control blocks it.

Open, Configurable, and Conditional Chaining Logic

Build your own reusable attack path logic from scratch, drawing on any event or action from your threat arsenal: TTPs, payloads, custom actions, or library content. Chain actions together and branch them based on conditions (if a credential is valid, pivot here; if a control blocks the step, stop or reroute) so scenarios adapt in real time to what’s actually in the environment instead of running a fixed script. This logic stays fully inspectable and configurable; update it anytime as new techniques emerge, and because OpenAEV is part of Filigran’s XTM Platform, it can be built and prioritized around the threat intelligence relevant to your organization.

Live, Transparent, Actionable Attack Path Graph Mapping

Watch the simulation unfold in real time on an interactive attack path graph, tracking every hop, pivot, and branch as it happens, from first action to final objective, instead of a static report after the fact. Every action’s result (a credential, an IP/port, a token, a file, and more) is captured as a structured finding, and the engine evaluates it at runtime to decide the next step. Drill into any finding to see exactly why and how an action fired, and use the graph to spot chokepoints—the single step whose removal breaks the entire path—so teams can fix one control instead of triaging every weakness the chain touches.

Operator vs. Agent-Led Orchestration

Attack Chaining can be run manually, fully orchestrated by AI agents end-to-end, or as a combination of both with a human-in-the-loop, all on the same conditional chaining engine and scope controls. In operator-led mode, you build the logic and control execution step by step, deterministic and transparent, best for scaling pentesting and validating attack paths with precision. In agent-led mode, you set the objective and scope in plain language and the agent builds, drives, and adapts the chain itself, reacting to findings as they surface and generating phishing emails and landing pages for social-engineering steps, best for automated pentesting and red teaming at scale.

“A validation outcome is only actionable when security teams can trace the logic that generated it,” said Jean-Philippe Salles, VP of Product Management at Filigran. “With OpenAEV v3, teams can build or generate attack scenarios, watch attack paths unfold, and inspect the logic and evidence behind every step. That means they can test more often, adapt scenarios to their own environment, and identify the specific control or weakness that needs attention without turning every validation exercise into a custom engagement.”

Also in OpenAEV v3

The release introduces a redesigned home dashboard that consolidates posture, simulation results, and detection coverage in a single view, known as the Adversarial Exposure Command Center. There’s also an Adversarial Exposure Score that aggregates validation results across exposure sources into one trackable measure; native AI red-teaming injectors that run adversary simulations against LLM-powered agents and chatbots using the same engine that validates EDR, SIEM, and email defenses; and one-click PDF reporting generated directly from simulation results.

Availability

OpenAEV v3 is available to all users today. Attack Chaining is available in OpenAEV Enterprise Edition.

Read more about Attack Chaining at the Filigran blog.

About Filigran

Filigran, a cybersecurity company, offers an open-source, AI-powered, threat-informed approach to Continuous Threat Exposure Management (CTEM). Its eXtended Threat Management (XTM) platform delivers threat intelligence, exposure validation, and cyber risk reduction. Learn more: Website – Blog – LinkedIn – X

FAQs

How is automated attack path validation different from a penetration test?

A penetration test depends on booking scarce specialist time against a fixed schedule, so most teams get a point-in-time picture once or twice a year, and the environment starts drifting from it the moment the report lands. OpenAEV runs the same multi-step reasoning on demand and in parallel across environments, which does not replace the judgment a human tester brings but does let teams validate continuously and reserve that expertise for the chokepoints the automation surfaces.

How does OpenAEV Attack Chaining use AI?

Attack Chaining utilizes an agentic layer for fully autonomous, end-to-end orchestration, empowering the AI to dynamically plan, execute, and adapt attack paths on the fly with minimal human intervention. Crucially, every decision the agent makes is logged against strict scope controls, ensuring teams get the realism of an automated red-teamer while keeping the entire exercise completely auditable.

Will an automated attack simulation break something in production?

Every chain runs inside scope controls that define which systems it may touch and how far it may travel, and execution halts on its own when a control blocks a step instead of forcing past it. This is why OpenAEV chains can run unattended in production-like environments rather than requiring a supervised window.

How is OpenAEV’s attack path validation different from other attack-path or automated pentesting tools?

Most tools either replay predefined scenarios or expose only the final result, while OpenAEV decides each step from what the previous one returned and lets teams inspect and modify that logic at any point. The same engine also runs against LLM-powered agents and chatbots, so teams can validate their AI systems with the tooling they already use for EDR, SIEM, and email defenses.

Media gallery

About The Author
Business Wire