Ethical Artificial Intelligence Statement
A Formal Commitment to Responsible Technology in Senior Safety
Effective January 1, 2026
Preamble
Beacon Boxx, LLC (“Beacon Boxx,” “we,” or “the Company”) develops and operates a safety monitoring and emergency response platform purpose-built for adults aged 55 and older residing in active adult communities, independent living facilities, and senior housing environments. Our platform leverages artificial intelligence (“AI”) and data-driven technologies to detect emergencies, monitor resident safety, and connect communities with timely, accurate response capabilities.
We recognize that the individuals we serve — older adults and the community operators, caregivers, and family members who support them — place profound trust in us. That trust is not incidental to our business; it is the foundation of it. We further recognize that AI systems, when deployed without deliberate ethical governance, carry real risks: the potential to erode privacy, amplify bias, reduce human agency, and cause harm to vulnerable populations.
This Ethical AI Statement (“Statement”) constitutes Beacon Boxx’s formal organizational commitment to the responsible design, deployment, and continuous governance of AI within our platform and operations. It applies to all AI and machine learning systems used or developed by Beacon Boxx, all employees, contractors, and third-party partners involved in the design or operation of those systems, and all products and services offered under the Beacon Boxx brand.
1. Our Ethnical AI Principles
Beacon Boxx’s approach to AI is governed by six foundational principles. These principles are operational commitments that carry specific obligations, and against which our AI systems and practices will be regularly measured.
Principle 1: Human Safety Above All
Every AI capability deployed by Beacon Boxx exists for one overriding purpose: to protect human life and well-being. No commercial interest, efficiency gain, or technological objective shall ever take precedence over the physical safety, emotional security, or dignity of the residents and communities we serve. When a system’s output could directly affect safety outcomes — including fall detection, emergency alerting, or anomaly detection — human review and intervention pathways shall always be preserved. AI augments human judgment in life- safety contexts; it never replaces it.
Principle 2: Privacy as a Right, Not a Feature
The residents served by Beacon Boxx are entitled to robust privacy protections as a matter of fundamental right. The data generated by their daily lives is deeply personal. Beacon Boxx commits to:
Collecting only the data strictly necessary for the delivery of safety and emergency response services;
Never selling, licensing, or commercializing resident data to third parties for advertising, profiling, or any non-safety purpose;
Providing clear, plain-language disclosure to residents and community operators regarding what data is collected, how it is used, how long it is retained, and with whom it is shared;
Implementing privacy by design principles across all product development;
Complying with all applicable federal and state privacy laws, including HIPAA where applicable.
UPDATED v2.0:
Health & Biometric Data (from Apple HealthKit):
Heart rate (real-time and resting), heart rate variability (HRV), blood oxygen saturation (SpO2)
Breathing rate, VO2 Max, step count, sleep analysis (sleep stages, duration, schedule)
Blood pressure (where user has a paired connected device)
Electrocardiogram (ECG) readings (where user has Apple Watch Series 4+ with ECG authorized)
Fall detection events and AFib detection alerts
Body temperature (where supported by device)
Clinical Health Records
(from Apple HealthKit Clinical Records — NEW in v2.0):
Prescription medications: medication name, dosage, and frequency as logged in the user’s Apple Health record. This data is collected with explicit user authorization and is used solely to populate emergency alert messages transmitted to first responders and designated emergency contacts.
Diagnosed medical conditions: health conditions logged in the user’s Apple Health clinical record. This data is collected with explicit user authorization and is used solely for the same emergency alert and first-responder notification purposes.
Both medication and medical condition data are treated as sensitive health information and are subject to the highest level of data protection applied by Beacon Boxx. This data is never used for advertising, profiling, predictive analytics unrelated to safety, or shared with any third party other than designated emergency contacts and first responders as part of an active emergency response.
Apple Watch as Independent Data Endpoint (NEW in v2.0):
As of this version, the Beacon Boxx platform has been architected to allow the Apple Watch to function as an independent, authenticated data endpoint. The Watch maintains its own authenticated session with the Beacon Boxx backend using credentials stored in the Watch Keychain, separate from and independent of the user’s iPhone session.
This architectural change means:
The Apple Watch transmits sensor data, alert events, and aggregated health vitals directly to the Beacon Boxx backend without requiring the iPhone to be active, in range, or authenticated;
Monitoring and emergency detection can continue while the user’s iPhone is idle, charging, or out of
Bluetooth range;
The Watch operates as a first-class data source with its own authentication lifecycle, token refresh
mechanism, and data transmission pathway;
Data transmitted directly from Watch to backend is subject to the same privacy protections, retention
policies, and security standards as data transmitted via the iPhone.
Residents will be informed of the Watch’s independent data transmission capability during the app’s Quick Start Wizard upon first login, and in the app’s plain-language privacy disclosure. Users can disable Watch-independent monitoring from within the app settings at any time.
Principle 3: Fairness and Freedom from Bias
Principle 4: Transparency and Explainability
Residents, family members, community operators, and care professionals deserve to understand how Beacon Boxx’s AI systems work. We commit to providing accessible, non-technical explanations of how our systems function; clearly disclosing the role of AI in any alert, recommendation, or decision delivered through our platform; maintaining internal documentation available for regulatory review; and never deploying AI systems so opaque as to preclude meaningful human oversight of life-safety functions.
Principle 5: Human Dignity and Autonomy
The adults served by Beacon Boxx retain the right to make meaningful choices about their own care, safety, and privacy. We commit to designing AI features with the input of older adult users; ensuring residents retain meaningful control over their personal data including the right to access, correct, and request deletion; never deploying AI-driven monitoring in ways that are covert or stigmatizing; and providing accessible opt-out mechanisms for non-safety-critical features.
Principle 6: Accountability and Continuous Governance
Ethical AI is a practice, not a declaration. Beacon Boxx accepts organizational accountability for the outcomes produced by our AI systems. We commit to designating an AI Ethics and Compliance function; conducting annual AI ethics audits; maintaining accessible reporting channels for AI ethics concerns; investigating concerns in good faith; and publicly reporting on AI governance practices annually.
2. Scope of Application
UPDATED v2.0:
Section 2 updated in v2.0 to reflect expanded data categories and Watch independent data endpoint.
This Statement applies in full to:
All AI and machine learning systems developed, licensed, or operated by Beacon Boxx, including fall
detection algorithms, emergency alert classification models, behavioral anomaly detection systems,
predictive safety analytics, and any AI-assisted communication or engagement features;
All health and biometric data collected through Apple HealthKit, including the new clinical record
categories (prescription medications and diagnosed medical conditions) added in Version 2.0;
All data transmitted by Apple Watch operating as an independent authenticated data endpoint, as
described in Principle 2 above;
All employees of Beacon Boxx involved in the design, development, deployment, maintenance, or
evaluation of AI systems;
All third-party vendors, contractors, and development partners — including offshore development firms — who develop or contribute to AI systems used in Beacon Boxx’s platform;
All community operators and partner organizations who access Beacon Boxx’s AI-powered features
through our platform.
Third-party partners are required to acknowledge and operate in a manner consistent with this Statement as a condition of their engagement with Beacon Boxx.
3. Prohibited Uses of AI
Beacon Boxx expressly prohibits the following uses of AI within its platform, operations, and partner ecosystem:
Any use of AI to make fully automated, non-reviewable decisions regarding the physical safety, medical assessment, or emergency response protocols for individual residents without a human-in-the-loop mechanism;
The use of resident data — including prescription medications and medical conditions collected via
HealthKit clinical records — to train AI models for purposes unrelated to safety, emergency response, or the direct delivery of Beacon Boxx’s contracted services;
The deployment of facial recognition, continuous audio surveillance, or other highly intrusive monitoring technologies without explicit, informed, and documented consent;
The sale, transfer, or licensing of resident behavioral data, clinical health records, safety event histories, or AI-derived profiles to third parties for commercial, research, or governmental purposes without explicit legal authorization and resident consent;
The use of AI to generate, infer, or communicate medical diagnoses, clinical assessments, or treatment
recommendations, which remain exclusively within the scope of licensed healthcare professionals;
Any application of AI that knowingly produces discriminatory outcomes or that is designed to
disadvantage, manipulate, or deceive the residents, families, or communities served by Beacon Boxx.
4. Governance and Implementation
4.1 AI Ethics and Compliance Function
Beacon Boxx shall maintain a designated AI Ethics and Compliance function responsible for: overseeing adherence to this Statement; conducting and coordinating periodic AI ethics audits; reviewing and approving deployment of new AI capabilities; and serving as the primary contact for internal and external AI ethics concerns.
4.2 AI Development Standards
All AI systems deployed by Beacon Boxx or developed by third parties on our behalf shall be subject to a documented development and validation process including: requirements documentation specifying ethical design objectives; bias testing against representative population subsets; validation against defined safety thresholds prior to deployment; and ongoing post-deployment monitoring with defined escalation triggers.
4.3 Incident Response
In the event that an AI system produces an incorrect, harmful, or potentially biased output affecting resident safety or well-being, Beacon Boxx commits to: promptly investigating the root cause; notifying affected parties transparently; implementing remediation measures; and documenting the incident and response for internal review and, where appropriate, external disclosure.
4.4 Training and Awareness
Beacon Boxx shall provide regular training to all relevant employees on the principles of ethical AI, the requirements of this Statement, and their individual obligations. Training shall be updated as this Statement evolves and as the AI landscape develops.
4.5 Review and Amendment
This Statement shall be reviewed no less than annually. Amendments may be made to reflect changes in Beacon Boxx’s technology, applicable law, regulatory guidance, or industry best practices. Material amendments shall be publicly disclosed. This Version 2.0 constitutes such a material amendment and supersedes Version 1.0 in its entirety.