AI Terms for the Wilma platform

These AI Terms apply to Wilma platform features that use machine learning, generative AI, or large language models to assist with recruitment workflows, including transcription, summarisation, drafting, and structured workflow outputs.

The Wilma platform may use AI systems to support features such as:

• Transcription of audio or video content
• Summaries of candidate responses or application materials
• Structured extraction of information from candidate submissions
• Email drafting and editing assistance for recruiters
• Workflow support for scheduling and candidate communication

These outputs are generated to assist human recruiters and hiring teams using the Wilma platform. They are not intended to replace human review or judgment.

AI outputs in Wilma are assistive only. Customers remain fully responsible for reviewing outputs, verifying their accuracy against the underlying candidate materials and job requirements, and making all employment-related decisions.

Wilma may process customer and candidate content through third-party AI providers solely to provide requested product functionality for the Wilma platform. This may include sending candidate-submitted application materials, transcripts, recruiter instructions, job criteria, and related workflow context to AI providers for inference and output generation.

This processing is used to generate outputs for the relevant customer workflow. It is not used by Wilma to train general-purpose AI models unless explicit permission is obtained.

Where third-party AI providers are used, Wilma seeks contractual safeguards restricting the provider’s use of customer and candidate data, including restrictions on using that data to train general-purpose models unless separately authorised.

Customers should not submit special category data, highly sensitive data, or data they are not authorised to process through Wilma’s AI features.

Wilma may process customer and candidate content through third-party AI providers solely to provide requested product functionality for the Wilma platform. This may include candidate-submitted application materials, transcripts, recruiter instructions, job criteria, employer context, and related workflow data.

Wilma does not use customer or candidate data to train general-purpose AI models unless explicit permission is obtained.

Customers may not use Wilma’s AI features to:

• Make fully automated employment decisions without meaningful human review
• Reject, advance, rank, or score candidates solely on the basis of AI outputs
• Infer or score protected characteristics or sensitive traits
• Assess emotion, facial appearance, attractiveness, accent, voice quality, or perceived confidence
• Create hidden suitability, culture-fit, personality, or behavioural scores from AI outputs
• Generate discriminatory or unlawful hiring criteria
• Process content they do not have a lawful basis to use

AI features may change, be improved, or be withdrawn over time. Some AI-assisted features may be experimental, especially during early access or design partner phases.

Questions about Wilma’s AI features can be sent to privacy@withwilma.com.