Freely subscribe to our NEWSLETTER

People search sites (PSS) are specialized data brokers that collect and sell personal data, which can include home addresses, phone numbers, and property records, among other details. Often, people are completely unaware that such data about themselves is accessible and searchable.

Incogni’s analysis reveals that 75% of House members’ personal data can be found on these sites. This data may include information such as home addresses, phone numbers, relatives, property records, and even neighbors.

The safety of all US politicians is under the spotlight following the recent attempted assassination of former President Donald Trump. The House of Representatives’ top security official has urged members of Congress to take advantage of all available security resources as they go back to their districts during recess1. Meanwhile, a man from Virginia has been charged with making threats of violence against Vice President Kamala Harris.

According to the study, male Republican House members face the highest risk, with 79.79% of their records being found on these sites compared to 73.98% for male Democrat House members. Female Republican House members have a slightly lower exposure rate of 61.76%, but this remains a significant security concern.

All House members from Arkansas, Georgia, Iowa, Michigan, Nevada, Oregon, Tennessee, and Utah have their data exposed on PSS with a probability of 75% or greater (see Methodology). The least-exposed data pertains to House members from Connecticut, New York, Indiana, Minnesota, Washington, Arizona, and South Carolina—personal information for less than 60% of them can be found on people search sites.

Darius Belejevas, data privacy expert and head of Incogni: “The safety of US lawmakers has rarely felt so precarious, so it was frightening to uncover how much of their personal data is easily obtainable.

This is a massive oversight that exposes House Members to security risks and needs to be urgently addressed.

This isn’t just an issue affecting high-profile politicians, though. Millions of Americans can be found on people search sites and their data can be exploited by fraudsters and identity thieves or sold to unknown entities for purposes they haven’t consented to.”

Methodology

Incogni’s researchers examined what data is available online regarding US politicians—specifically members of the House. The sample was derived on July 15, 2024, and included 437 representatives, including non-voting delegates and excluding three vacant seats.

Incogni checked dozens of people search sites for each member in the sample using publicly available data: each House member’s full name, date of birth, and, where available, their current state and city of residence. This initial personal data was found either on the government website congress.gov/members or Ballotpedia. If a state or city of residence was unavailable, we used the representative’s birthplace.

We then checked dozens of people search sites for each member in our sample, using this publicly available data to perform the searches.

For this study, we chose 75% as the match probability at which we deemed a member of the House discoverable on people search sites. For every individual data point we used to query the PSS—name, surname, age, city and state of residence—a match ranking in the 1 – 100 range was assigned. In cases where personal details were found to be a partial match, such as a slight difference in age or spelling variations, the match ranking for that data point fell below 100. The algorithm determines a match ranking for every data point based on the magnitude of the difference between a query and its potential match. The match rankings for every data point are then averaged and presented as an overall match ranking, expressed as a percentage.

Manually checking a random sample of results with the threshold probability of 75% (we checked 15/172 such results), we found that 93% were, in fact, matches.

Given that we observed numerous correct matches at lower probabilities (6/15 at 70%, for example), we believe that potential false positives are compensated for by uncounted matches (those with a probability below the 75% threshold).