1. Introduction
TheSHFTApp LLC ("TheSHFTApp," "we," "us," or "our") is committed to developing and deploying AI technology responsibly, ethically, and transparently. This Responsible AI Policy sets forth the principles, standards, and governance frameworks that guide our use of AI Systems across the Platform.
This Policy applies to all AI Systems operated by TheSHFTApp, whether built internally or provided through third-party AI model providers, and governs AI development, deployment, monitoring, and improvement practices.
This Policy should be read alongside our AI Disclosure, Privacy Policy, Acceptable Use Policy, and Terms of Use.
2. Core Principles
TheSHFTApp's responsible AI framework is grounded in the following principles:
2.1 Transparency. We are committed to disclosing how AI Systems are used in the Platform through our AI Disclosure and this Policy. We will notify users when they are interacting with AI-generated content where material.
2.2 Human Oversight. AI Systems on the Platform are designed to assist and augment human decision-making, not to replace it. No AI System makes final determinations about hiring, employment, placement, or material outcomes affecting users without appropriate human involvement or review.
2.3 Fairness and Non-Discrimination. We work to identify and mitigate bias in our AI Systems. We do not design AI features intended to discriminate against individuals based on protected characteristics including race, color, sex, national origin, religion, age, disability, or other legally protected categories.
2.4 Privacy by Design. AI Systems are developed with privacy considerations embedded from design through deployment. Data minimization, purpose limitation, and user rights are incorporated into AI system architecture. See our Privacy Policy.
2.5 Safety. We implement safety measures to prevent AI Systems from producing harmful, abusive, deceptive, or dangerous content. We maintain abuse detection, content filtering, and safety review processes.
2.6 Accountability. TheSHFTApp takes responsibility for the AI Systems it deploys. We establish clear internal ownership, monitoring, and governance for AI features.
2.7 Continuous Improvement. We treat responsible AI as an ongoing commitment, not a one-time compliance exercise. We iterate on our AI practices as technology, regulation, and best practices evolve.
3. AI Governance Framework
3.1 AI Review Process. New AI features and significant changes to existing AI Systems undergo internal review before deployment, including assessment of potential risks, bias, data handling, and user impact.
3.2 Monitoring. Deployed AI Systems are monitored for performance, safety, accuracy, and abuse. Anomalies or quality issues trigger internal review and remediation.
3.3 Third-Party AI Providers. We evaluate AI model providers for their safety practices, data handling commitments, bias mitigation efforts, and reliability before integration. We contractually require appropriate standards from providers.
3.4 Internal Responsibility. Designated personnel are responsible for AI system oversight, incident response, and regulatory compliance related to AI features.
4. Bias and Fairness
4.1 Bias Assessment. We conduct assessments of AI Systems to identify potential disparate impact or bias across demographic groups, particularly for features that affect user opportunities such as career matching and candidate scoring.
4.2 Mitigation Measures. Where bias is identified, we take corrective action, which may include retraining models, adjusting weighting, expanding training data diversity, or modifying feature design.
4.3 No Guarantees. We cannot represent that our AI Systems are entirely free from bias. We commit to good-faith, ongoing effort to identify and reduce bias.
4.4 Employment Law Considerations. AI features used in talent screening, candidate ranking, or employer-facing matching are designed with awareness of applicable anti-discrimination laws. However, users (particularly Employers) bear independent responsibility for ensuring their use of Platform features complies with applicable employment law.
5. AI Content Safety
5.1 Prohibited AI Outputs. Our AI Systems are configured to avoid generating content that:
- Is harmful, threatening, or abusive;
- Facilitates illegal activity;
- Is sexually explicit or exploitative;
- Promotes discrimination or harassment;
- Is materially false or deceptive; or
- Violates applicable law.
5.2 Safety Filters. We implement content moderation, safety classifiers, and prompt safety systems to reduce the generation of harmful outputs.
5.3 Jailbreaking and Circumvention. Users are prohibited from attempting to bypass, disable, or circumvent AI safety controls. Such conduct is a violation of our Acceptable Use Policy and may result in account suspension.
5.4 User Reporting. Users may report problematic AI Outputs by contacting support@theshftapp.com. We investigate all reports and use them to improve our safety systems.
6. AI and Employment Decisions
6.1 AI as an Assistive Tool. AI Systems on the Platform are designed to assist job seekers with career preparation and to assist Employers and Recruiters with candidate discovery. AI Systems do not make final employment decisions.
6.2 Employer Responsibility. Employers and Recruiters using AI-powered candidate matching, scoring, or profiling features are solely responsible for ensuring their hiring practices comply with applicable employment discrimination laws and any applicable regulations regarding automated employment decision tools.
6.3 Candidate Rights. Candidates may request information about how AI-generated profiles or scores may have been used in connection with their candidacy by contacting privacy@theshftapp.com.
7. Data Use for AI Improvement
7.1 De-identified Data. We may use De-identified, aggregated data from AI interactions to evaluate, improve, and fine-tune AI Systems, subject to privacy safeguards.
7.2 Embeddings and RAG. We may maintain vector embeddings and retrieval-augmented generation systems using processed forms of user data to enable context-aware AI responses.
7.3 User Controls. Where we provide opt-out mechanisms for AI training use of personal data, those preferences are honored as described in the Privacy Policy.
7.4 No Unconsented Third-Party Training. We will not share individually identifiable user data with third-party AI model providers for the purpose of training their general models without user consent.
8. Responsible AI in Employment Technology
TheSHFTApp is aware of evolving regulatory frameworks related to AI in employment decision-making, including the EU AI Act, proposed U.S. federal guidance, and various state and local laws regarding automated employment decision tools. We monitor these developments and intend to update our practices and disclosures to comply with applicable requirements as they take effect.
9. Generative AI and Authenticity
9.1 User Disclosure Responsibility. Users who submit AI-generated content (such as AI-generated resumes or cover letters) to employers, job boards, or third parties are solely responsible for complying with any disclosure requirements or restrictions imposed by those third parties.
9.2 AI-Assisted Does Not Mean Inauthentic. AI-assisted content is a legitimate writing and editing aid. However, AI-generated content must accurately represent the user's actual qualifications, experience, and credentials. Submitting AI-generated content that misrepresents material facts is a violation of the Acceptable Use Policy.
10. Policy Updates
As AI technology and applicable regulations evolve, this Policy will be updated. Material updates will be communicated consistent with our Terms of Use amendment procedures.
11. Contact
For questions about our responsible AI practices: Email: legal@theshftapp.com Support: support@theshftapp.com
See also: AI Disclosure, Privacy Policy, Acceptable Use Policy.