AI and Responsible Hiring Policy
1. Purpose
This Policy explains how SHIFTLY L.L.C-FZ ("Shiftly") designs and governs artificial intelligence in the Platform. AI supports human decision-making in recruitment workflows. It does not replace human judgement and does not operate as an unsupervised decision-maker.
2. What we use
The Platform uses automated matching and ranking. It uses AI to assist with CV parsing, profile analysis and analytical insights, and offers optional assessment and interview functionality provided through specialist services. It also uses automated signals for security and fraud detection. Not every feature is enabled for every Employer or Talent, and features in use are indicated in the product.
3. Decision-support and human oversight
These features are decision-support tools. Shiftly does not use AI to make hiring decisions with legal or similarly significant effect on individuals, and does not decide on its own whom an Employer must hire or reject. Human oversight is required for any decision informed by these outputs. An Employer may override a ranking or recommendation, a Talent is not rejected from an employment opportunity solely because of an AI-generated recommendation, and Employers are expected to review the underlying evidence before deciding. Shiftly identifies AI-generated or AI-derived insights where needed so that Users understand their nature and limitations. Employers and Institutions are responsible for their hiring decisions and for their own compliance with applicable law.
4. Structured eligibility rules
The Platform may apply structured eligibility or workflow rules that an Employer configures, such as a factual requirement or a screening question. A Talent may be marked ineligible on the basis of such a rule or answer. These are structured Employer criteria, not AI decisions, and the Employer is responsible for defining and applying them lawfully.
5. Employer responsibilities
An Employer that uses the Platform's functionality to evaluate candidates remains responsible for its use of that functionality and for its hiring decisions. The Employer is responsible for the requirements that apply to its use of automated tools where it hires, including any notice, consent or bias-audit obligation. Additional requirements may apply where AI functionality is used in another jurisdiction, and Shiftly may add notices, controls or regional terms before making particular functionality available there.
6. Prohibited practices
Shiftly does not use AI to identify people through facial biometrics, create voiceprints, carry out biometric identification or categorisation, infer race, ethnicity, religion, sexual orientation, health, political beliefs or other protected characteristics, score physical attractiveness, detect lies, infer personality from facial expressions, or perform emotion recognition prohibited by applicable law. This prohibition is not lifted because a User has consented to a related feature. Consenting to a recorded interview does not authorise emotion inference from the recording. Ordinary recording, transcription and analysis of the content of interview answers, and psychometric assessment, remain permitted subject to disclosure.
7. Accuracy and bias
Automated and AI outputs may be incomplete, inaccurate or affected by bias in data or methods. Match recommendations and explanations are informational and should not be treated as objective truth or a guarantee. Shiftly does not guarantee that its features are unbiased or accurate. Shiftly applies reasonable safeguards intended to identify materially anomalous, discriminatory or inappropriate outputs, keeps the human-oversight requirement in clause 3, and reviews reported concerns.
8. Training and our no-training position
Shiftly does not use identifiable candidate data, including CVs, profiles, application materials, interview outputs, messages or an Employer's confidential candidate data, to train or fine-tune AI models shared with other customers or provided to third parties. Where model development or testing is needed, Shiftly prefers synthetic, licensed or genuinely anonymised data. This position concerns Shiftly's own use. A specialist service that Shiftly integrates is governed by its own terms and by Shiftly's arrangements with it, and Shiftly does not make commitments on that service's behalf beyond what those terms support.
9. Transparency and correction
Shiftly discloses its use of AI through this Policy, the Privacy Policy, the notice shown before an AI-assisted interview, and the product itself. You may ask Shiftly to correct a factual error in your profile or in data derived about you by contacting privacy@shiftly.ae. Contesting an Employer's hiring decision is a separate matter that Shiftly does not decide. Shiftly's route concerns the accuracy of the underlying data, not the merits of an Employer's decision.
10. Providers and changes
Shiftly may integrate specialist AI services under appropriate contractual and data-protection obligations. Because these services and the law change quickly, Shiftly may replace a service, or modify, restrict or withdraw a feature, where necessary for legal, security, quality or availability reasons, and will give business Customers reasonable notice where a change materially affects a paid feature.
11. Aggregated data
Shiftly may use genuinely anonymised or aggregated data, from which individuals cannot reasonably be re-identified, for analytics, benchmarking, service improvement and security.
12. Liability and contact
Liability from the use of these features is governed by the limitation-of-liability clause in the Terms and Conditions, which this Policy does not vary. Questions about this Policy: privacy@shiftly.ae.