Terms & Privacy

Purpose: educational and academic use

This site exists as a learning project and a public demonstration of techniques for working with civic open-data feeds. It is not a commercial product, not a journalistic outlet of record, and not affiliated with any government, university, or news organization. The code, methodology, and outputs are intended for study, discussion, and civic curiosity — not as an authoritative source.

No guarantee of accuracy

Public Analyst.ai aggregates public crime-incident data from each city's open-data portal and normalizes it across jurisdictions with differing schemas, category definitions, reporting practices, and historical migrations (e.g. UCR to NIBRS). Despite best efforts to map and reconcile these sources faithfully — documented on the methodology page — we make no warranty, express or implied, about the accuracy, completeness, timeliness, or fitness for any particular purpose of any number, chart, signal, forecast, or piece of prose published on this site.

Source datasets contain known issues: late-reported incidents, reclassifications, redacted locations for sensitive categories, and changes in police-department reporting policy. Our normalization layer can introduce additional error. The site is provided "as is" and "as available," without warranty of any kind.

Data sources

Incident data is pulled from each city's official open-data feed (DataSF, City of Chicago Data Portal, NYC Open Data, LA City open data, City of Oakland open data, Seattle Open Data, etc.). Population figures for per-capita rates come from the U.S. Census Bureau ACS 5-year estimates. Neighborhood boundaries follow each city's official analytical units. All sources are public; none of the data on this site is private, proprietary, or obtained through restricted channels.

Source data is the property of its respective publisher and is used here under the terms of the relevant open-data license. Public Analyst.ai is not affiliated with, endorsed by, or operated on behalf of any city, police department, or government agency.

Forecasts and statistical signals

Twelve-month forecasts on this site are generated using Prophet, an open-source time-series forecasting library released by Meta under the MIT license. Anomaly signals (spikes, drops, sustained shifts, rare events, streak breaks) are computed from rolling baselines using the rules described on the methodology page.

Forecasts are statistical projections, not predictions. They reflect patterns in past data and assume conditions broadly similar to the recent past — they do not account for policy changes, demographic shifts, economic shocks, or any other future event. Backtest performance is published per category and per city; please consult those figures before relying on any specific forecast.

Not for safety, legal, or commercial decisions

Nothing on Public Analyst.ai constitutes legal, financial, real-estate, insurance, employment, or public-safety advice. The site is an educational and academic exercise intended for learning, research, and civic curiosity. Do not use it as the sole basis for:

• real-estate purchase, rental, or valuation decisions;
• insurance underwriting or pricing;
• employment, lending, or tenant screening;
• policy or law-enforcement deployment decisions;
• any decision affecting personal safety or legal status.

Crime data describes reported incidents, which is not the same as actual crime, risk, or neighborhood quality. Reporting rates vary by category, neighborhood, and over time.

Limitation of liability

To the maximum extent permitted by law, Public Analyst.ai and its operators shall not be liable for any direct, indirect, incidental, consequential, special, or exemplary damages arising from use of, or inability to use, this site or its data — including but not limited to errors, omissions, inaccuracies, downtime, or decisions made in reliance on the content.

Privacy

Public Analyst.ai does not require accounts and does not collect personally identifiable information from visitors. No incident data on the site identifies individual victims, suspects, or witnesses; published incidents are aggregated to the neighborhood × category × month level, and the underlying public datasets themselves redact or generalize sensitive fields per each city's open-data policy.

The site may use standard web analytics and hosting infrastructure (e.g. Vercel) which collect aggregated, non-identifying request metadata such as IP address, user-agent, and referrer for the purpose of operating and securing the service. We do not sell, share, or use this metadata for advertising or profiling.

Intellectual property

Original prose, charts, signal definitions, and page layouts on Public Analyst.ai are © 2026 Public Analyst.ai. Underlying public data remains the property of its source publishers. You are welcome to cite, quote, and link to the site with attribution; please do not scrape, mirror, or republish full pages without permission.

Changes to these terms

These terms may be updated as the site evolves. Material changes will be reflected in the "Last updated" date at the top of this page. Continued use of the site after changes constitutes acceptance of the revised terms.