A Security Delivers End-to-End Autonomous Offensive Security with Launch of Continuous Internal Infrastructure Testing

Platform now autonomously discovers, remediates and validates attack paths from the perimeter to crown jewel data, reducing risk across the entire attack surface

NEW YORK, Nov. 12, 2026 (GLOBE NEWSWIRE) -- A Security, the autonomous offensive security and remediation platform, today extended its autonomous testing, remediation and exploit validation capabilities from the perimeter into the network and private cloud with the launch of Continuous Internal Infrastructure Testing. The new capability enables organizations to autonomously assess how an AI-powered adversary would exploit their infrastructure from within, measure the business impact of the potential compromise and address the identified attack chains at machine speed while continuously validating the fixes. This gives security teams end-to-end coverage of AI-powered attack paths and protection of crown jewel data, from the outside in, within a single platform.

Adversaries are actively searching for and gaining access to enterprise environments through an exposed server, vulnerable application or stolen credentials. Organizations that test external exposures and internal networks separately mistakenly overlook how a low-severity vulnerability and a misconfigured account together can yield a path to full network control. Threat actors with a valid account and map of internal infrastructure simply route around modern defenses, such as EDR, MFA or network segmentation, by chaining together weaknesses that no single control was designed to see. Combined with A’s continuous testing of external infrastructure and applications, the new Continuous Internal Infrastructure Testing enables security teams to test the paths and chained attacks threat actors exploit from the perimeter into the internal network.

“Weaponized AI lets attackers move through enterprise networks faster than ever by adapting to defenses and linking together weaknesses that are typically addressed in isolation,” said Omer Gull, Chief Product Officer, A Security. “We already help Fortune 1000 organizations find and close the external exposures attackers use to get in, and with Continuous Internal Infrastructure Testing, we now cover everything from the perimeter to the crown jewels within a single platform. Customers asked us to show what an attacker could reach from a single compromised account, fix the weaknesses that make those paths possible and keep proving they stay closed as their environments change. Every path we close takes risk off the table and puts critical data further out of an attacker’s reach.”

As AI-driven attack capabilities become more accessible, threat actors are using autonomous agents to break down defenses, change tactics and pursue alternative routes to critical systems. Recent research revealed the China-nexus group VAPOR PANDA used AI agents to run a cyber espionage campaign targeting more than 200 organizations across at least 14 countries, with the agents conducting reconnaissance, harvesting credentials and moving laterally while switching attack paths when an approach failed. Unlike scripted tools that stop when an initial attempt is blocked, agents continuously adjust their approach until they find a path in. Capable open-weight models that are available for download and require no verification or content filters put that persistence within reach of any attacker at a fraction of the cost. Continuous Internal Infrastructure Testing allows security teams to outpace attackers by identifying how those weaknesses can create viable attack paths, remediating them and consistently validating the fixes as environments and attack techniques evolve.

Key capabilities A Security's Continuous Internal Infrastructure Testing include:

• Autonomous Internal Attack Testing: Continuous Internal Infrastructure Testing starts from a single domain user account, the kind gained through phishing or stolen credentials. A Security’s agents then map the network and private cloud environment and pursue objectives the customer sets. They adapt as they go and test new vulnerabilities as soon as they’re disclosed.
  
  
• Attack Chain Discovery, Remediation and Validation: Agents link vulnerabilities and misconfigurations, including those without a CVE, into evidence-backed chains mapped to MITRE ATT&CK. These chains show how low-severity issues add up to a full takeover. A Security then autonomously remediates each chain and keeps retesting it to ensure it remains closed.
  
  
• From the Perimeter to the Internal Network on One Platform: Internal testing now runs continuously alongside A Security’s external and application testing. This gives security teams one view of how attackers get in and how they move through the network and across private cloud infrastructure. Lightweight connectors reach remote and segmented sites, customers set the scope and approvals and Layered Guardrails prevent destructive actions.
  
  
• Bring Your Own Cloud or Model: For regulated and air-gapped environments, organizations can deploy A Security in their own AWS, Azure or Google Cloud account and route AI calls only to approved models. Findings, evidence and credentials stay within boundaries they already govern, and no internet access is needed.
  
  

Continuous Internal Infrastructure Testing is available through an Early Access Program. For more information, and to apply, visit go.a.security/Internal-Infrastructure-Early-Access.

About A Security
A Security is the autonomous offensive security and remediation platform built for organizations where failure is not an option. With layered guardrails built in, the platform runs the full offensive lifecycle, continuously finding and chaining vulnerabilities to validate and eliminate real attack paths before weaponized AI can exploit them. The platform gets sharper with every campaign, logs every action with full auditability, and breaks the economics of attacks by demonstrating and remediating the weaknesses that malicious agents will exploit. Founded in 2025, A Security is backed by Lightspeed and Cyberstarts.

Media Contact
Geoff Lopes
Marketbridge for A
A@marketbridge.com