Eagleview Artificial Intelligence Transparency and Data Disclosure

Summary of Datasets Used in Model Development 

This document provides information for California consumers, pursuant to California Civil Code Section 3110 et seq., on how we, Eagle View Technologies, Inc., and our affiliates and subsidiaries (“Eagleview”), use data to train Eagleview’s generative artificial intelligence services.: 

Dataset Sources: All aerial imagery is collected and owned by Eagleview. No third-party imagery sources are used for training. 

Intended Purpose: Datasets are used to train algorithms that detect and describe features for geospatial and property analytics. 

Number of Data Points: Billions of aerial image tiles representing petabytes of data across urban and suburban U.S. regions. 

Types of Data Points in Datasets: High-resolution imagery, geospatial metadata (coordinates, altitude, timestamps), and human annotations of property and structure features. 

Inclusion of Public Domain Or IP-Protected Data: No third-party imagery sources are used for training. 

Acquisition of Datasets: No datasets are purchased or licensed from third parties for the use of training Eagleview’s algorithms. 

Inclusion of Personal Information Or Aggregate Consumer Information: Datasets do not knowingly include personal information as defined in Cal. Civ. Code §1798.140(v). Imagery is collected from public airspace and is not collected for identification; though it may incidentally capture objects/structures visible from public airspace; not used for consumer profiling. No consumer profiles, behavioral data, or aggregate consumer information are used. 

Cleaning, Processing, And Other Modification to Datasets: Images undergo orthorectification, color correction, and quality review for consistency and accuracy. Human annotations are used for training computer vision models. 

Data Collection Time Period: Imagery collection is ongoing. Historical datasets span more than ten years with continuous updates. 

Dates Datasets First Used: Machine learning, deep learning, large vision model development, and other AI development began as early as 2020 and continues to the present day and use only Eagleview primary data and annotations applied to those data by Eagleview employees. 

Use of Synthetic Data: Eagleview does not currently, knowingly use ‘synthetic data generation’ as defined in Cal. Civ. Code §3110(e) in development of its production models.