A plain-language description of our AI systems, written for candidates whose applications are reviewed with Aigis and for the employers who use it. Last updated: July 2026.
If an employer told you they use Aigis
An employer reviewing your application may use Aigis to help their team read and compare CVs. This page explains exactly what the AI does with your application, what it cannot do, and what rights you have. It exists so that “we use AI-assisted screening” is never a black box.
What the screening AI does
- The employer defines assessment criteria for the role (skills, experience, qualifications), usually extracted from the job description and then reviewed and edited by them.
- The AI reads each CV and gives each criterion a match level (Strong / Solid / Partial / Gap) with a confidence level and — crucially — quoted evidence from the CV itself justifying that assessment. Employers see words and evidence, not a numeric score to rubber-stamp.
- Quotes are verified against the actual CV text by code: if the AI cites something that isn't really there, the evidence is downgraded to unverified and the confidence is reduced. The AI is not allowed to grade on vibes.
- Criteria the employer marks as essential can raise a flag when they appear unmet. A flag is a prompt for human attention — it does not reject anyone.
What the AI cannot do
- It cannot reject you. There is no automated rejection path in Aigis. Moving a candidate to “rejected” always requires an explicit human action, and employers agree in our terms to maintain meaningful human review of every outcome.
- It does not learn from your data. Your CV is never used to train AI models.
- It does not keep your data indefinitely. CV data is automatically and permanently deleted 28 days after last activity, with a hard maximum of 180 days — including the file, extracted text, all AI assessments, and search index entries.
- It does not share your data. Your CV is visible only to the hiring team that uploaded it, is encrypted at rest, and is never sold.
The AI Twin
Separately from screening, candidates can create an AI Twin on Aigis — a conversational AI trained only on documents the candidate themselves uploads and controls. Anyone who chats with a Twin is told they are interacting with an AI system. Twins answer only from the candidate's own content, cite their sources, and say “I don't have evidence for that” rather than inventing answers.
Your rights as a candidate
- You have the right to know AI is used in a recruitment process that affects you (the employer must tell you).
- You have the right not to be rejected solely by automated processing (GDPR Article 22), to ask for human intervention, and to contest an outcome.
- You can ask the employer what data of yours was processed, ask for it to be corrected or deleted, and object to AI-assisted processing.
- If you can't reach the employer or believe something is wrong, contact us at hello@aigis.bio and we will act on verified requests.
For employers
If you screen candidates with Aigis, you are the data controller for the CVs you upload, and you must tell candidates that AI-assisted screening is used — you are welcome to link to this page in your job adverts and candidate communications. A copy-paste notice template and a practical checklist are in our Responsible AI Screening guide.
Regulatory context
AI systems used to evaluate job candidates are classed as high-risk under the EU AI Act, whose transparency rules apply from 2 August 2026 and whose full high-risk framework applies from 2 December 2027. Aigis is designed around those requirements — AI-interaction disclosure, evidence-cited output, human-in-the-loop by construction, logging, and strict retention — and GDPR applies to all candidate data processing today. See our Privacy Policy and Terms of Service.