Cybersecurity leaders can now use a new framework to find and address weaknesses that could leave their AI systems vulnerable to ransomware attacks.
The tool is the Ransomware Assessment and Incident Defense Framework for AI Systems (RAID-AI) and is fully available for anyone to use online for free.
Behnaz Karimi and Yuvaraj Govindarajulu, both co-leads of the Open Worldwide Application Security Project’s (OWASP) AI Bill of Materials (AIBOM) initiative developed RAID-AI over the past two years.
They had initially planned to introduce the framework at the ISC2 Security Congress, held in Denver, Colorado, between October 26 and 28. Unfortunately, visa issues prevented Karimi from attending.
A Manual Ransomware Readiness Assessment for AI Systems
Speaking to Infosecurity, Karimi said RAID AI focuses on ransomware attacks against AI systems, whether or not an attacker uses AI. The framework’s approach is hands-on, asking security teams to assess their own controls rather than connecting to or scanning their systems.
“Right now, we are slowly seeing that ransomware attacks targeting AI systems are happening,” Karimi said.
She highlighted a April 2025, a report by Zoran, Mitrovic head of engineering at Kintra Health, which detailed how ransomware actors used data poisoning techniques, as well as an alleged cyber-attack on Novo Nordisk in June 2026, where hackers claimed to have stolen AI models used by the pharmaceutical company.
The RAID-AI assessment covers general cybersecurity measures alongside risks specific to AI models, data and supporting infrastructure.
The framework is delivered through a web console aimed at CISOs, cybersecurity officers and security engineers.
Users assess controls manually, selecting maturity levels on a scale from zero to five. The controls are organized into general and AI-specific categories, with levels such as basic, advanced and high.
General controls include recovery, training and sensitive-data protection. AI-specific areas include data, tools and runtime.
RAID AI maps controls to two voluntary security frameworks, the US National Institute of Standards and Technology’s (NIST) Cybersecurity Framework 2.0 (CSF 2.0) and the US Cybersecurity and Infrastructure Security Agency’s (CISA) Cybersecurity Performance Goals 2.0 (CPG 2.0).
It also allows users to map adversarial attacks through three attack mapping frameworks, MITRE ATT&CK, MITRE ATLAS and NIST Adversarial.
Karimi and Govindarajulu have provided comprehensive documentation, including guidelines for using the tool and an explanation of how it was developed and how maturity levels and various security scores are calculated.
Future Plans: Extend Ransomware Assessments to Agentic AI Systems
Organizations do not need to connect their internal systems or upload company documents to complete the assessment.
Therefore, the RAID-AI assessment is not an automated audit: results depend on users’ evaluations and the tool does not independently verify an organization’s security posture.
Instead, teams review their practices and enter their assessments themselves.
“Ideally bringing security managers, engineers and AI-security specialists together in workshops,” Karimi told Infosecurity.
The console displays the results in dashboards, including maturity views and mappings to other frameworks. Users can also download PDF reports that highlight weaknesses and help teams set priorities.
Karimi said the dashboard can give senior leaders a high-level overview, while technical teams use the underlying controls to guide more detailed discussions.
In the future, Karimi hopes to provide practical implementation resources, including videos and blueprints to help security engineers apply individual controls, and develop a similar tool for ransomware threats targeting agentic AI systems.