Ethical AI

    AI Ethics & Fairness Guidelines

    Last updated: February 10, 2026

    Our Commitment to Ethical AI

    At Hirebyai, we believe artificial intelligence should enhance human decision-making, not replace it. We are committed to building AI systems that are fair, transparent, and accountable. Our AI ethics framework guides every aspect of our platform design and operation.

    This document outlines our principles, practices, and ongoing commitments to responsible AI in recruitment. We hold ourselves accountable to these standards and continuously work to improve our systems.

    Core Principle: AI assists, humans decide. Our platform provides insights and recommendations, but all final hiring decisions are made by people, not algorithms.

    1. Human-in-the-Loop Decision Making

    We design our AI systems to augment human judgment, not automate it away:

    • No Automated Rejections: Our AI never automatically rejects candidates. All decisions require human review
    • Explainable Recommendations: Match scores include clear reasoning (e.g., "skills match: 85%", "experience level: suitable")
    • Override Capability: Recruiters can override AI recommendations with their professional judgment
    • Human Validation: Critical decisions (final hiring, rejection) always involve human assessment

    Why This Matters: AI can identify patterns but lacks the contextual understanding, empathy, and judgment that humans bring to hiring decisions. We use AI to surface insights, not to make choices.

    2. Bias Detection and Mitigation

    We actively work to identify and reduce bias in our AI systems:

    Protected Characteristics Redaction

    Before AI processing, we redact or de-emphasize information that could lead to bias:

    • Names (to reduce name-based bias)
    • Age and date of birth
    • Gender identifiers
    • Addresses (to reduce location-based bias)
    • Photos and profile pictures
    • Nationality and immigration status (where not legally required)

    Skills-First Matching

    Our AI focuses on:

    • Demonstrated skills and competencies
    • Relevant experience (not just years)
    • Transferable skills from different industries
    • Career progression and growth potential

    Regular Bias Audits

    We conduct quarterly audits to check for:

    • Disparate impact across demographic groups
    • Systematic patterns that could indicate bias
    • Feedback from users about unfair outcomes
    • Model performance across different candidate profiles

    3. Transparency and Explainability

    We believe users have the right to understand how AI affects their opportunities:

    For Job Seekers

    • Match scores include clear breakdowns (skills, experience, preferences)
    • CV analysis highlights matched and missing keywords
    • Skill gap analysis shows what to improve
    • Rejection feedback (when provided by employers) is shared transparently

    For Employers

    • Candidate rankings include reasoning (why this match score?)
    • Filtering criteria are clearly displayed
    • AI suggestions are labeled as "suggestions", not "requirements"
    • Employers can see which factors influenced each match

    No Black Boxes: We don't use opaque "trust the algorithm" approaches. Every AI decision is accompanied by reasoning that users can understand and evaluate.

    4. Fairness by Design

    Fairness is built into our platform architecture, not added as an afterthought:

    • Diverse Training Data: We use diverse datasets to train our models, avoiding over-representation of any single demographic
    • Fairness Metrics: We track equal opportunity, demographic parity, and disparate impact metrics
    • Accessibility: Our platform meets WCAG 2.1 AA standards for accessibility
    • Multiple Pathways: Candidates can showcase skills through CVs, portfolios, projects, and endorsements. Not just traditional credentials
    • Skill-Based Search: Employers can search by skills and competencies, reducing reliance on proxies like university prestige

    5. Data Privacy and PII Protection

    Ethical AI requires responsible data handling:

    • PII Redaction: Personal identifiable information is redacted before CVs are processed by our proprietary AI engine
    • Encryption: AES-256-GCM encryption for all sensitive data at rest and TLS 1.3 in transit
    • Data Minimization: We collect only the data necessary to provide our services
    • Automated Deletion: CV data is automatically deleted after 90 days; users can request deletion anytime
    • No Data Selling: We never sell candidate or employer data to third parties

    See our Privacy Policy for full details on data handling.

    6. Legal and Regulatory Compliance

    We comply with all applicable UK and EU regulations:

    • UK GDPR & EU GDPR: Full compliance with data protection regulations
    • Equality Act 2010: Our platform supports employers in meeting non-discrimination requirements
    • ICO Guidelines: We follow Information Commissioner's Office guidance on AI and data processing
    • Employment Law: Our features align with UK employment law best practices
    • EU AI Act Readiness: Monitoring upcoming EU AI Act requirements for high-risk AI systems

    7. Accountability and Continuous Improvement

    We hold ourselves accountable for our AI systems' impact:

    Feedback Mechanisms

    • Users can report unfair outcomes or biased results
    • Dedicated AI ethics review process for reported issues
    • Quarterly transparency reports on bias audits and platform metrics

    Continuous Monitoring

    • Real-time monitoring of AI model performance and fairness metrics
    • A/B testing to ensure changes don't introduce bias
    • Regular retraining on updated, diverse datasets

    Independent Review

    • Third-party audits of our AI systems (planned for 2026)
    • Collaboration with academic researchers on bias detection
    • Transparency in publishing findings and improvements

    8. Known Limitations and Ongoing Challenges

    We acknowledge that no AI system is perfect. Current challenges we're actively working on:

    • Career Gaps: Our AI may undervalue non-traditional career paths or career breaks (e.g., parental leave, caregiving)
    • Skill Transferability: AI struggles to fully recognize transferable skills from non-IT backgrounds
    • Regional Variations: UK-wide salary data may not capture hyper-local variations
    • Emerging Roles: New job titles and emerging technologies may not have sufficient training data

    We actively research and develop solutions to these challenges. User feedback is critical to identifying and addressing limitations.

    9. Your Rights Regarding AI Decisions

    Under UK GDPR, you have rights related to automated decision-making:

    • Right to Human Review: Request human review of any AI-generated recommendation
    • Right to Explanation: Ask for detailed reasoning behind match scores or recommendations
    • Right to Contest: Challenge unfair AI outcomes and receive a response
    • Right to Opt-Out: Request that your profile not use AI matching (manual review only)

    To exercise these rights, contact us at contact@agniaisolutions.com. We will respond within 30 days.

    Questions or Concerns?

    We welcome feedback, questions, and concerns about our AI ethics practices:

    AI Ethics Team: contact@agniaisolutions.com

    Report Bias: contact@agniaisolutions.com

    General Support: contact@agniaisolutions.com

    We're Listening: Your feedback directly influences how we improve our AI systems. Every report is reviewed by our AI ethics team.