Delivering trustworthy AI models require simple and easy tools for model testing, validation and documentation that can meet business and regulatory requirements. Leveraging its 25 years of experience in Risk Management, CIMCON has created AI Validator, an automated test suite that provides a comprehensive range of AI model tests ranging from fairness, interpretability and validity to code quality, privacy and cybersecurity to meet stringent regulatory requirements. AI Validator automatically identifies high risk models by assigning a risk score to each model.
Benefits
- Cut time spent on Model Development, Testing and Validation
- Automated Risk Scoring, Code Quality and Model Dependencies
- No-code Inventory and Workflow Management
- Evidence controls and compliance through automated test results and supporting documentation
- Scan models/data for privacy and vulnerabilities
- Aligned with NIST AI Framework
Features
Code Quality
- The Code Quality helps users to know the details of errors and warnings generated for the AI file.
Link Map
- The Link Map helps users to understand the link of the AI file with libraries, inputs & outputs. It helps to show CVE for the AI libraries.
Vulnerability
- The Vulnerability Test helps users to find CVE for the AI libraries.
Privacy
- The Privacy Test helps users to search for privacy-related keywords within AI file.
Risk Score Card
- The Risk Score Card allows the users to analyze and review the Attributes, Attribute Type, Attribute Count, Attribute Risk Factor associated with the AI file and their Risk Contribution.
AI Content Detection
- The Generative AI Content/AI Model helps users to find if the content/model is based on the AI or not.
LLM Risk Assessment
- LLM Risk Assessment identifies vulnerabilities in LLM-generated responses to ensure reliable and trustworthy content.
LLM Hallucination
- The LLM Hallucination Test helps users understand the Hallucination Rate of custom LLMs trained on a specific dataset.
LLM Source Attribution
- The LLM Source Attribution test helps users understand where with a structured & unstructured data an LLM is generating a response to a prompt from.
Fairness
- The Fairness Test evaluates the fairness of AI models. It assesses whether the predictions or decisions made by the AI model exhibit bias or discrimination towards certain groups based on sensitive attributes such as race, gender, or age.
Interpretability
- The Interpretability Test helps users to find the importance of input features in the model's decision-making process, highlighting which features have the most significant impact on the predictions.
Validity & Reliability
- The Validity & Reliability Test helps users to evaluate the trustworthiness and robustness of AI models.
Why CIMCON’s AI Validator Module?
The use of spreadsheets, models, Access databases, Python, R, SQL, SAS scripts, RPA scripts, and other File tools continues to proliferate rapidly. File errors can lead to business losses, regulatory penalties and reputational damages that can run into hundreds of millions of dollars. File Insight Discovery scans your network to automatically create an inventory of high-risk files, identify errors, and generate an enterprise data lineage map.
Quickly gain high-level insight into your files with a fast scan
EUCI Discovery scans networks drives, user drives, and document management systems (SharePoint, Documentum, Box, Office 365, etc.) to give you reports on your file types, age, and size.
Identify errors & PII/PCI in files with automated risk assessment
EUCI Discovery further scans your files for complexity, and whether or not they contain any errors or sensitive information. It can also integrate with data classification applications to report on files that have or have not been tagged.
Gain visual insight into enterprise data flows & lineage
Links between files are common but these links are a risk to your business, if they are broken, stale, point to the wrong location, and need updates. EUCI Discovery scans for all links (upstream & downstream) and provides a visualization of linked files. Reports can also identify files with links that will break if migrated to cloud storage.
File Disposition Service (Add-On Feature)
After scanning your files, you may want to disposition them as well to get rid of millions , or even billions of files that are outdated and do not require active storage or management. Unneeded files continue to consume valuable I.T. infrastructure, effort and costs in storage, virus protection, backups and performance. EUC Insight offers an optional File Disposition service that performs automated dispositioning of files based on a rule-based engine/policies that you can define. For more information, click here.
Link Fixer Service (Add-On Feature)
When moving files to the Cloud, links to other files and data sources can often break thus invalidating the file outputs. EUC Insight offers an optional Link Fixer Service that can automatically detect these links, and then fix them either before or after migration. For more information, click here.
Features
Fast Scan
- Scan millions of files per hour for a high-level overview of your file landscape.
Analysis
- Identify files with errors, broken links, and hidden information.
Automated Risk Assessment
- Configurable risk algorithm based on complexity, materiality and other attributes.
Policy Compliance
- Make policy compliance easier with user-defined fields and a built-in criticality assessment form.
Enterprise Lineage Map
- FileLink™ technology visually displays how files are linked to each other & to enterprise databases.
GDPR & PII
- Detect credit card or social security numbers and other sensitive information (PCI,PII,PHI).
Content Migration
- Before Moving Excel Files to cloud storage, scan them to identify the links that will break and cause errors.
VBA/Macro Search
- Find hard coded passwords and other sensitive keywords in VBA code and macros.
AI Risk Management Framework
Explore the realm of Artificial Intelligence (AI) with our AI Risk Management Policy. This concise guide covers the spectrum of AI models, including supervised, unsupervised, and deep learning, and emphasizes making AI trustworthy based on the NIST AI Risk Management Framework.
Learn to assess and manage AI Risk, cultivate a culture of risk awareness, and utilize periodic testing with tools like ours. This policy is your essential toolkit for responsible and effective AI utilization in your organization.