How to Protect Your Privacy From Predictive Policing

Police departments across the country are increasingly using artificial intelligence to predict where crimes will happen and who might commit them. Known as predictive policing, this practice feeds vast amounts of data — including records purchased from data brokers, social media activity, and location histories — into algorithms that generate risk scores and crime forecasts. While proponents argue it makes policing more efficient, civil liberties organizations warn that it creates a surveillance infrastructure that threatens the privacy and constitutional rights of millions of innocent people.

What Is Predictive Policing?

Predictive policing is the use of data analytics, machine learning, and AI algorithms to forecast criminal activity. It generally falls into two categories:

• Place-based prediction: Algorithms analyze historical crime data to identify geographic areas where crimes are statistically more likely to occur, generating "hot spot" maps that direct patrol officers to specific neighborhoods or blocks
• Person-based prediction: Algorithms assign risk scores to individuals based on factors like arrest history, social connections, geographic proximity to past crimes, and data from commercial sources — flagging them as likely future offenders or victims

Major predictive policing platforms include PredPol (now Geolitica), HunchLab, Palantir Gotham, and various proprietary tools built by companies like RELX and Thomson Reuters. By 2025, the technology was being used by police departments in dozens of U.S. cities, including Los Angeles, Chicago, New York, and Atlanta.

What Data Does It Use?

Predictive policing systems are data-hungry. They ingest information from a wide range of sources, many of which most people never realize are being used for law enforcement purposes:

• Historical crime reports and arrest records: The foundational dataset for most systems, including offense type, location, time, and suspect demographics
• Data broker records: Companies like RELX and Thomson Reuters provide police with access to vast databases of personal information — addresses, phone numbers, associates, financial records, and more — often without a warrant
• Social media monitoring: Posts, connections, and activity on platforms like Facebook, Instagram, and X can be scraped and analyzed for threat indicators
• Location data: Cell phone location records, automatic license plate reader (ALPR) data, and GPS information purchased from commercial data brokers
• Surveillance footage: Camera networks, including those equipped with facial recognition, feed into some predictive systems
• ShotSpotter/gunshot detection: Acoustic sensors that detect gunfire and automatically alert police to specific locations
• 911 call data and field contact reports: Records of police interactions that may not have resulted in arrests

Why Predictive Policing Is a Privacy Threat

The civil liberties concerns surrounding predictive policing are substantial and well-documented.

Mass Surveillance of Innocent People

Predictive policing does not target specific suspects based on probable cause. Instead, it casts a wide net over entire communities, analyzing the data of millions of people who have never been accused of a crime. When your social media posts, location data, and personal records are fed into an algorithm that generates a risk score, you are effectively being surveilled and judged without your knowledge or consent.

Racial Bias and Discrimination

Multiple studies and investigations have found that predictive policing algorithms perpetuate and amplify existing racial biases in the criminal justice system. Because these systems are trained on historical crime data — data that reflects decades of disproportionate policing in Black and Latino communities — they tend to direct even more police attention to those same neighborhoods. The NAACP has documented how this creates a feedback loop: over-policing generates more arrests, which generates more data, which the algorithm interprets as higher risk, which leads to even more policing.

Lack of Transparency and Accountability

Most predictive policing systems are proprietary, meaning the algorithms are trade secrets shielded from public scrutiny. According to research from Rutgers AI Ethics Lab, most American police departments lack clear policies on algorithmic decision-making and provide little to no disclosure about how these systems work. Communities cannot challenge what they cannot see, and individuals who are flagged as high-risk may never know it happened or have an opportunity to contest the designation.

Constitutional Concerns

The use of predictive policing raises serious Fourth Amendment questions. Purchasing data from brokers to circumvent warrant requirements, using geofencing and IMSI catchers to collect data on innocent people en masse, and making policing decisions based on algorithmic predictions rather than individualized suspicion all push against fundamental constitutional protections against unreasonable searches.

How to Protect Your Privacy From Predictive Policing

While systemic change requires policy reform and public advocacy, there are practical steps you can take right now to reduce the amount of personal data available to predictive policing systems.

1. Remove Your Data From Broker Sites

Data brokers are one of the primary pipelines feeding personal information into predictive policing systems. Companies like RELX and Thomson Reuters sell law enforcement access to databases containing your name, address history, phone numbers, email addresses, known associates, financial information, and more. Removing your data from these sources directly reduces what is available to police algorithms.

PrivacyOn automates data removal from 100+ data broker sites, including the people-search platforms that aggregate and sell your personal information. With 24/7 continuous monitoring and re-removal as brokers re-collect your data, PrivacyOn helps keep your information out of the databases that feed surveillance systems. Family plans cover up to 5 people starting at $8.33/month.

2. Limit Your Location Data Exposure

Location data is a key input for predictive policing. Minimize your exposure by:

• Disabling location history on your phone (Google Timeline, Apple Significant Locations)
• Setting app permissions to "While Using" or "Never" instead of "Always" — most apps do not need constant location access
• Using a VPN to mask your IP-based location from websites and online services
• Avoiding free apps that monetize your data by selling location information to brokers
• Paying with cash when possible, since card transactions create a location trail

3. Tighten Your Social Media Privacy

Social media monitoring is a documented component of predictive policing systems. Protect yourself by:

• Setting all social media profiles to private
• Removing or limiting personal information in your bios (location, workplace, school)
• Being cautious about posting real-time location updates, check-ins, or geotagged photos
• Reviewing and removing old posts that contain personal details
• Limiting your connections to people you actually know

4. Understand Your Rights

Know your legal protections and exercise them:

• FOIA requests: You can file Freedom of Information Act requests to learn whether your local police department uses predictive policing tools and what data sources they rely on
• State privacy laws: Some states have enacted consumer privacy laws that give you the right to know what data companies collect about you and to request deletion
• Right to protest: Support and participate in local advocacy for transparency and accountability in police use of surveillance technology

5. Reduce Your Overall Digital Footprint

The less personal data that exists about you online, the less material predictive policing algorithms have to work with:

• Use privacy-focused browsers and search engines (Firefox, Brave, DuckDuckGo)
• Opt out of data collection wherever possible — loyalty programs, warranty registrations, free Wi-Fi sign-ups
• Use encrypted messaging apps like Signal for sensitive communications
• Regularly audit and delete unused online accounts
• Consider using a privacy-focused mobile OS like GrapheneOS for maximum control

6. Monitor Your Exposure With Dark Web Scanning

Personal data that has been exposed in breaches often ends up in law enforcement databases and commercial data broker records. PrivacyOn includes dark web monitoring that alerts you when your personal information appears in criminal marketplaces or leaked datasets, giving you the opportunity to respond before the data spreads further.

The Bigger Picture

Predictive policing is part of a broader trend toward algorithmic governance, where AI systems make or influence decisions that profoundly affect people's lives — from who gets policed more heavily to who gets hired, insured, or approved for a loan. Protecting your privacy from these systems requires both individual action and collective advocacy for transparency, accountability, and meaningful regulation.

The data broker industry is the connective tissue that makes much of this surveillance possible. By removing your personal information from broker databases, limiting your digital footprint, and staying informed about how your data is being used, you can significantly reduce your exposure to predictive policing and the broader surveillance ecosystem it represents.

Take control of your personal data. Start protecting your privacy with PrivacyOn today.