AI Transparency Policy

1. How AI Is Used

SkillDrill uses AI in the following areas:

• Skill conversations: AI conducts adaptive, natural-language conversations with team members to identify skills, experience, and competencies. The AI follows conversation prompts defined by your organisation and asks intelligent follow-up questions.
• Skill extraction: AI analyses conversation responses to identify and categorise skills mentioned by participants, including proficiency levels.
• Team analysis: AI generates insights about team composition, strengths, gaps, and recommendations.
• Professional concern detection: If a team member discloses a professional concern during a conversation (e.g. workload issues, organisational problems), the AI flags it for review by the appropriate person. SkillDrill does not store or process medical or health data; only professional disclosures made in the context of work.
• Prompt generation: AI assists administrators in creating effective conversation prompts from question set topics.
• Scenario analysis: AI answers hypothetical workforce questions based on your organisation's skill data.

2. Data Processing

When AI processes your data:

• Conversation data is sent to our AI model provider via encrypted API calls
• Data is processed in real time and not stored by the AI provider beyond the request
• Your data is never used to train AI models
• All data is encrypted in transit (TLS 1.2+) and at rest
• Each organisation's data is isolated using tenant-specific encryption keys

3. Human Oversight

AI in SkillDrill is designed to assist and augment human decision-making, not replace it:

• AI-generated skill profiles and analyses are always presented as insights, not definitive assessments
• Administrators can review, edit, and override any AI-generated content
• Professional concerns flagged by AI are routed to designated reviewers for assessment and action
• Question prompts generated by AI can be reviewed and modified by administrators before use
• Team members can review and correct their own skill profiles

4. Bias & Fairness

We are committed to minimising bias in our AI systems:

• Our conversation prompts are designed to be neutral and inclusive
• AI skill extraction focuses on objective skills and experience, not demographic characteristics
• We regularly review AI outputs for patterns that may indicate bias
• Organisations can customise question sets to reflect their specific context and values
• We welcome feedback on potential bias issues and investigate all reports

5. Data Security

Protecting your data in AI processing is a top priority:

• All AI API calls use encrypted connections (TLS 1.2+)
• Personal data (emails, names) is encrypted at rest using tenant-specific keys
• AI providers are bound by data processing agreements that prohibit data retention and model training
• We use enterprise-grade AI API endpoints with enhanced privacy protections
• Audit logs track all AI interactions for accountability

6. Opt-Out

Participation in AI-facilitated conversations is voluntary. Team members can:

• Choose not to participate in skill conversations
• Request that their conversation data be deleted
• Manually update their skill profile without AI interaction
• Contact their organisation administrator to discuss alternatives

Organisation administrators can control which features use AI and can disable AI-powered features at any time.

7. Updates

This policy is reviewed quarterly and updated to reflect changes in our AI capabilities, third-party providers, or relevant regulations. Material changes will be communicated to all active customers at least 30 days before taking effect.

We are committed to evolving our AI practices in line with emerging best practices, UK AI regulations, and the EU AI Act where applicable.