Skip to content

AI governance

How Club AI is governed for high-risk hiring under the EU AI Act — transparency, human oversight, and your data.

Last updated Questions trust@thequantumclub.com

How Club AI is governed

  1. Your data is never used to train models

    Candidate CVs, notes and meeting transcripts are never used to train or fine-tune public LLMs. Only internal, anonymised telemetry calibrates our own prompts.

  2. A human always decides

    Club AI is assistive. Match scores, summaries and recommendations inform people; they never make an automated hiring decision.

  3. Hiring AI is high-risk — we treat it that way

    The EU AI Act classifies recruitment AI as high-risk. We design for transparency (users know when AI is involved), human oversight, and against adverse impact.

  4. Assessments are screening-support, not verdicts

    Our assessments are positioned as observed work-sample evidence, never standalone selection or clinical/personality diagnosis.

  5. AI sub-processors are named

    The model providers behind Club AI are listed in our sub-processor table, with region and data-handling terms.

Regulation

Where we stand on the EU AI Act

EU AI ActAligned Transparency and human-oversight duties for the AI-assisted hiring features of Club AI, treated as a high-risk use under the Act. Our position on the EU AI Act

Model providers

Who processes AI-related data

As listed in the sub-processor register.

Vendor
AI model providers (Anthropic, Google)
Purpose
Club AI features (matching, assistant)
Data
Prompts — never used to train models
Region
EU / US (SCCs)
DPA
Yes
Listed since
The sub-processor register

From the security review FAQ

Questions on AI

Do you use our data to train AI models?
No. Customer data is never used to train or fine-tune public AI models. Prompts sent to model providers are processed under contract and not used for training.
How do you handle hiring AI under the EU AI Act?
Club AI is assistive and a human always makes the decision. We build for transparency, human oversight, and against adverse impact, and treat assessments as screening-support only.
How are your AI sub-processors governed?
The model providers behind Club AI (Anthropic, Google) are named in our sub-processor table with region and data terms, and each is bound by a DPA with SCCs. Prompts are processed under contract and never used to train or fine-tune their models. We list them so you always know which third parties touch AI-related data.