The Conformity Assessment Process (CAP)
Delivering Trusted, Safe, Fair and Responsible AI through systematic evaluation and verification.
What is CAP?
A systematic evaluation to verify AI systems meet safety, ethical, and performance standards. CAP ensures AI "works as intended."
The AI Dilemma
AI is transforming industries, but when it goes wrong, it goes really wrong. We need to ensure AI is safe, fair, and reliable.
Why CAP Was Created
To provide a structured way to check if AI products meet the right standards before they hit the market.
Based on Global Standards
Created from extensive research on international AI standards including EU AI Act, IEEE Ethics standards, and Kenya's Code of Practice.
Data Evaluation
Assessing dataset quality, diversity, and bias mitigation to ensure fair and representative AI systems.
Model Assessment
Evaluating AI models for fairness, accuracy, and security to prevent harmful outcomes.
Ongoing Monitoring
Ensuring sustained ethical and performance standards post-deployment through continuous evaluation.
Our Approach (CAP)
CAP is divided into 5 phases which address specific aspects of AI safety and responsibility. The process is adaptable to different AI system types, use cases and applications.
Ownership Verification
Confirming product legitimacy and responsible development
Data Evaluation
Assessing dataset quality, diversity, and bias mitigation
Model Assessment
Evaluating AI models for fairness, accuracy, and security
Validation & Testing
Running technical and functional compliance checks
Ongoing Monitoring
Ensuring sustained ethical and performance standards
- Intellectual Property
- Development Practices
- Documentation Review
- Data Quality
- Bias Detection
- Representation Analysis
- Performance Metrics
- Fairness Testing
- Security Evaluation
- Functional Testing
- Standards Compliance
- Edge Case Analysis
- Performance Tracking
- Drift Detection
- Continuous Improvement
EU AI Act
Comprehensive assessment aligned with European regulations
IEEE AI Ethics Standards
Ethical guidelines for responsible AI development
Submit Your AI for Assessment
Give people confidence in your AI through proper validation, certification, and ongoing monitoring.
- Verify compliance with global standards
- Identify and mitigate potential risks
- Build trust with your users and stakeholders
- Receive guidance for continuous improvement
Submit for CAP Assessment
Elevate your AI product's impact with verified quality and compliance.