Compliance

Flock’s regulatory compliance efforts encompass a broad range of activities aimed at adhering to federal and state laws and regulations, as well as industry standards. This includes complying with requirements from agencies like the FBI’s policies for Criminal Justice Systems (CJIS), the US Data Privacy Framework, and the NIST Cybersecurity Framework

Privacy by Design

Privacy by Design requires a proactive, ethical, and integrated approach for the deployment of principles in product architecture and technologies to enable privacy rights into Flock’s solutions from the start. Principles including data minimization, limitations on retention, and least privileged access facilitates privacy for customers and communities.

Responsible Use of AI

We recognize that the power of Machine Learning in cybersecurity comes with a profound responsibility. We believe that ethical AI is fundamental to building trust and ensuring our technology serves its intended purpose of enhancing safety and efficiency without compromise.For us, AI Ethics means rigorously upholding principles such as:

Fairness: We are dedicated to building ML models that perform equitably across diverse real-world conditions, minimizing unintended outcomes that could arise from training data or environmental factors. Our processes include continuous monitoring and validation to ensure accuracy and impartiality in our classifications and readings.

Transparency: We strive to understand how our ML models arrive at their predictions. This commitment to explainability ensures that our technology can be auditable, its decisions understood, and its performance continuously improved.

Accountability: We take full ownership of the outcomes generated by our AI systems. Our internal governance structures and expert teams ensure that our models are developed, tested, and maintained responsibly.

Reliability: Our dedication to quality extends to the resilience of our ML models. Through rigorous testing, continuous validation, and a focus on adversarial robustness, we ensure our systems deliver consistent and dependable results in challenging environments.

Data Privacy and Security: Recognizing that ML relies on vast datasets, we implement stringent data protection measures throughout the entire data lifecycle. This includes secure data handling, access controls, and adherence to privacy regulations, ensuring the data used to train and operate our models is always protected.

Data Retention

We understand that effective data retention is not merely a regulatory checkbox; it's a proactive strategy to ensure the integrity of our systems,  and preserve both the confidentiality and availability of your information. Flock’s retention policy requires retaining customer data for a default period of 30 days and also provides flexibility where our customers demand different retention periods.

Data Collection and Processing

Data Elements Chart

Data Flock Collects
Description
ALPR & Vehicle Intelligence Data
License Plate Number
The alphanumeric characters extracted from the license plate
License Plate Image
The image frame containing the detected license plate
Vehicle Image
A broader image of the vehicle, providing context including model and color
Time Stamp
Date and time of the license plate capture
Additional Vehicle Attributes
Make, color, type, plate type and state, any notable markers
iOS Platform & Application Usage Data
User Identifiers
Unique ID’s assigned to customers’ users within the platform
Use Access Logs
User Identifiers
Unique ID’s assigned to customers’ users within the platform
Performance Data
Application performance metrics
Operational & Account Management Data
Customer Account Information
-Names, e-mail addresses, phone numbers of primary contacts
-Billing Information
-Contract Details
Support & Communication Data
Records of customer service interactions, support tickets, emails.
Contractual Data
Details of service agreements, SLAs, legal terms
Audit & Compliance Data
Logs related to user access to the platform, system changes and security incidents.
Machine Learning Training & Improvement Data
Annotated Image Data
License plate images, vehicle images, with human-verified bounding box annotations and labels
Model Performance Metrics
Data on model accuracy, precision, recall, false positive/negative rates, especially broken down by challenging conditions such as low light, angle, blur.
Customer Feedback on Performance
Aggregated and analyzed data from customer-flagged incorrect reads or classifications.