> For the complete documentation index, see [llms.txt](https://ultrasafe.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ultrasafe.gitbook.io/docs/images-and-media.md). # UltraSafe AI Data Usage Policy ### Our Commitment to Safe, Intelligent, and Executable Multimodal AI At UltraSafe AI, we are dedicated to enabling safe, intelligent, and executable multimodal AI systems designed for critical environments. Our mission is to enhance human abilities and everyday life while ensuring robust personal and public safety. This data usage policy reflects our unwavering commitment to these principles. ### Vision and Mission * Enabling Safe, Intelligent and Executable Multimodal AI: UltraSafe AI designs systems for critical environments, utilizing Ultra-Safe Intelligence to ensure autonomous operations adhere strictly to safety and ethical guidelines. * Promoting Human Welfare: UltraSafe AI enhances human abilities and everyday life while ensuring robust personal and public safety. * Safeguarding Ethical Standards: Our AI systems integrate ethical decision-making processes that deeply reflect human values. * Prioritizing Privacy: We implement state-of-the-art measures to ensure data is used responsibly, with full user consent and fortified protections against unauthorized access. ### How We Use Your Data At UltraSafe AI, we prioritize your data privacy and security above all else. As of August 22, 2024, data sent to the UltraSafe AI API is never used to train or improve our models. Your data remains yours, period. This commitment ensures that your proprietary information and user data stay completely confidential, aligning with our vision of prioritizing privacy and promoting human welfare. To maintain the highest level of security and abuse prevention, API data may be retained in an encrypted form for up to 15 days, after which it is permanently deleted (unless otherwise required by law). For our enterprise customers with highly sensitive applications, we offer zero data retention options. With zero data retention, request and response bodies are not persisted to any logging mechanism and exist only in encrypted memory to serve the request. This stringent data policy applies to all UltraSafe AI services, including our specialized industry models and research models. ### Default Usage Policies by Endpoint | ENDPOINT | DATA USED FOR TRAINING | DEFAULT RETENTION | ELIGIBLE FOR ZERO RETENTION | | ----------------------------- | ---------------------- | ------------------------- | --------------------------- | | /v1/chat/completions | No | 15 days | Yes | | /v1/embeddings | No | 15 days | Yes | | /v1/specialized/healthcare | No | 7 days | Yes | | /v1/specialized/finance | No | 7 days | Yes | | /v1/specialized/legal | No | 7 days | Yes | | /v1/specialized/cybersecurity | No | 7 days | Yes | | /v1/image/analysis | No | 15 days | Yes | | /v1/document/analyzer | No | 15 days | Yes | | /v1/data/reranker | No | 15 days | Yes | | /v1/audio/transcriptions | No | Zero data retention | - | | /v1/audio/translations | No | Zero data retention | - | | /v1/fine-tuning/jobs | No | Until deleted by customer | Yes | | /v1/moderations | No | Zero data retention | - | ### Technology Behind Ultra-Safe & Super Intelligent Multimodal Executable AI * Executable AI Focus: UltraSafe AI utilizes Safe Expert Models enhanced with Ultra-Safe Intelligence for exceptionally precise and reliable autonomous operations. * Advanced Algorithms: Our Safe Intelligent Expert Models employ cutting-edge machine learning algorithms that prioritize safety, including advanced anomaly detection and ethical reinforcement learning. * Continuous Monitoring and Adaptation: Our Safe Expert Models enable continuous, real-time monitoring of AI behavior for immediate, precise adjustments. * Transparency and Accountability: We maintain detailed logs and ensure that all AI actions are transparent and accountable, facilitating thorough reviews and bolstering trust in AI systems. * Ethical Data Usage: Our Safe Expert Models enforce the highest standards of data privacy, emphasizing responsible data use and robust security measures. ### Ethical AI and Decision Making UltraSafe AI integrates ethical decision-making processes that deeply reflect human values into all our systems. Our AI algorithms incorporate ethical guidelines that meticulously reflect human values, cultural sensitivities, and legal standards, ensuring decisions are made with ethical integrity and adapted to complex social contexts. * All data processing adheres to strict ethical guidelines. * Ethical considerations are built into our model architectures. * Regular audits ensure compliance with our ethical standards. ### Security, Transparency, and Accountability #### Comprehensive Security Measures * SOC 2 Type II, ISO 27001, and HIPAA compliance * Data encryption at rest (AES-256) and in transit (TLS 1.3) * Regular third-party penetration testing and vulnerability assessments #### Continuous Monitoring and Adaptation Our Safe Expert Models enable continuous, real-time monitoring of AI behavior for immediate, precise adjustments. This ensures: * Proactive risk mitigation * Consistent performance in critical environments * Rapid response to any anomalies #### Transparency and Accountability We maintain detailed logs and ensure that all AI actions are transparent and accountable, facilitating thorough reviews and bolstering trust in AI systems. This includes: * Comprehensive audit trails * Explainable AI features for decision traceability * Regular reporting on system performance and safety metrics ### Enhancing Human Abilities Safely Our strict data policies contribute directly to the overall safety and reliability of UltraSafe AI's systems, especially in critical environments. By ensuring data integrity and privacy, we enable: * Enhanced safety features with advanced fail-safe mechanisms * Proactive prevention of accidents through intelligent risk detection * Reliable autonomous operations that adhere strictly to safety and ethical guidelines ### Accessibility and Inclusivity UltraSafe AI is designed for universal accessibility. Our data usage policies support this goal by: * Ensuring diverse and representative data handling practices * Supporting customized interfaces and adaptive responses * Enabling AI systems that break down barriers for individuals with disabilities ### Benefits to Mankind * Enhanced Safety Features: Our AI systems feature advanced fail-safe mechanisms that automatically detect and mitigate potential risks, enabling proactive prevention of accidents. * Ethical Decision-Making: Our AI algorithms incorporate ethical guidelines that meticulously reflect human values, cultural sensitivities, and legal standards, ensuring decisions are made with ethical integrity and adapted to complex social contexts. * Accessibility and Inclusivity: UltraSafe AI is designed for universal accessibility, leveraging sophisticated algorithms to break down barriers for individuals with disabilities and promoting inclusivity through customized interfaces and adaptive responses. ### Enterprise Privacy at UltraSafe AI Trust, privacy, and security are the foundations of our mission at UltraSafe AI. We're committed to uncompromising privacy and security for all our services, including EUS1, EUS Mega, EUS Flash, and our specialized industry models. #### Our Commitments Ownership: You own and control your data * We never train on your business data (data from any UltraSafe AI service) * You own your inputs and outputs (where allowed by law) * You control how long your data is retained (customizable for all services) Control: You decide who has access * Enterprise-level authentication through SAML SSO and multi-factor authentication * Fine-grained control over access and available features * Custom models are yours alone to use and are not shared with anyone else Security: Comprehensive compliance and beyond * We've been audited for SOC 2 Type II, ISO 27001, and HIPAA compliance * Data encryption at rest (AES-256) and in transit (TLS 1.3) * Regular third-party penetration testing and vulnerability assessments * Visit our Security Portal for detailed information about our advanced security measures Safety and Ethics: Beyond Compliance * Advanced fail-safe mechanisms automatically detect and mitigate potential risks * Ethical reinforcement learning ensures AI decisions align with human values * Continuous adaptation to complex social contexts and emerging ethical considerations At UltraSafe AI, we go beyond industry standards to ensure that your data remains secure, private, and under your control at all times. Our commitment to privacy, safety, and ethics is not just a policy—it's the core of our technology and business practices, designed to promote human welfare and safeguard ethical standards in the age of AI.