What is OSINT, and where are its limits?

OSINT has become a marketing term attached to everything from serious records research to running a name through a consumer people-search site. The distinction matters, because the second produces exactly the confident, wrong answers that cause problems later.

What sources does genuine OSINT use?

• Court records: civil, criminal, probate, bankruptcy, and their dockets
• Property records, deeds, assessor data and liens
• Corporate and UCC filings, registered agents, officers and members
• Professional and occupational licensing registries
• Publicly visible social and professional profiles
• News archives, obituaries, and web archives of pages since changed
• Regulatory filings, permits, and government publications

Why is identity resolution the hardest part?

Because records describe names, not people. In a state the size of Tennessee there are many people sharing a common name, and aggregated people-search products routinely merge two or three of them into a single confident profile. The result is a report attributing someone else's judgment, conviction or bankruptcy to your subject.

Resolving it requires cross-referencing identifiers — date of birth, address history, relatives, middle names, employment — and frequently pulling the actual court file rather than trusting an index entry. This is where an investigator's time actually goes, and it is the step automated products skip entirely.

How stale is the data?

More than most users assume. Address data in aggregated products frequently lags by months to years, because it derives from sources that update irregularly. Court indexes vary by county in how promptly they post. Corporate filings are current only to the last annual report. For a question like 'where does this person live now', OSINT produces candidates, not answers — which is why a locate is finished in the field rather than on a screen.

Where does OSINT stop being open source?

At the point where you need someone's permission, credentials, or a false identity to see it. Sending a connection request from a fabricated profile to view restricted content is not OSINT. Nor is logging into an account that is not yours, guessing credentials, or asking a mutual acquaintance to extract private material. Those methods create legal exposure and taint whatever they produce.

The test is simple: could any member of the public see this without deceiving anyone or logging in as someone they are not? If yes, it is open source. If no, the lawful route is process issued by counsel.

What is OSINT genuinely good for?

Scoping and direction. It cheaply establishes what exists, what is contested, which entities and properties are involved, and where the real questions are — which is exactly what should happen before anyone spends money on field hours. The failure mode is treating it as the finished product rather than the map. A report that is entirely OSINT, with nothing verified, is a list of leads presented as conclusions.