Open-source intelligence · OSINT

When to commission open-source intelligence

An open-source intelligence assignment is most useful when a business decision depends on information that is fragmented, inconsistent or difficult to place in context. A proposed commercial relationship, an unexplained change in a company’s public profile or conflicting accounts of a regulatory development could justify focused research. The starting point should be the uncertainty affecting the decision, rather than a general request to find everything available.

For Lo Presti, agreeing the intelligence requirement establishes what the research must answer and where its limits lie. A commission might ask whether public evidence supports a stated ownership structure, or whether reported events materially alter an assessment of a business relationship. Defining the intended decision helps distinguish consequential findings from background information and prevents collection from becoming an end in itself.

Defining research questions and choosing sources

Precise questions give public-source research a defensible scope. Instead of asking whether a business is trustworthy, an assignment could examine which legal entities appear in its published materials, whether stated appointments agree with relevant filings, and how its public account of a particular event has changed. Jurisdictions, relevant dates and the entities under review should be settled before these questions guide collection.

Source selection follows the question. Corporate registers may help establish recorded ownership or appointments; regulatory notices may clarify an official action; public statements and local reporting may add chronology and context. Each has limits. A filing records information submitted within a particular system, while a news report may depend on unnamed sources. Neither should automatically be treated as a complete account of the underlying circumstances.

From collection to corroborated findings

Lo Presti’s approach evaluates information for authenticity, timing, relevance and independence. In practice, that means examining who published an item, what it actually establishes and whether later material changes its meaning. Repetition is not necessarily corroboration: several articles may reproduce one announcement, creating an impression of agreement without supplying separate evidence. Material conclusions need scrutiny beyond the number of search results supporting them.

Entity matching and chronology help test apparent connections. Similar company names, shared addresses or overlapping dates may warrant examination, but do not alone establish common control or a substantive relationship. Translations and automated analysis can assist with handling material; human review remains necessary for significant findings. Where sources conflict, the assessment should explain the disagreement and its implications rather than conceal it within a single confident narrative.

What an OSINT report should contain

A useful report makes the route from evidence to judgment visible. Lo Presti’s outputs can include a briefing supported by sources, a map of entities and relationships, an event chronology or a monitoring assessment. The appropriate format depends on the question: an ownership query may benefit from an entity map, while disputed descriptions of successive events may be clearer in a chronology.

Reporting should separate established facts, analytical interpretations and unresolved points. References should allow the reader to understand the origin and date of material evidence, with qualifications attached to the findings they affect. Coverage gaps also matter: an unavailable record or limited local reporting may restrict what can reasonably be concluded. The decision-maker should leave with a clearer understanding of both the findings and the strength of their evidential basis.

Hypothetical use: assessing a commercial relationship

Consider a hypothetical business reviewing a prospective distribution partner whose website describes a regional group, while available filings identify several similarly named companies. The research question would not be whether the group looks credible online. It would be which entities are identifiable, what public records establish about their relationships, and whether the entity proposed for the agreement matches the commercial description.

Research could compare filings, published company details and dated announcements, then present the relationships that can be evidenced alongside those that remain uncertain. A shared trading name might explain some ambiguity without resolving legal ownership. Practically, the findings could help the commissioning business identify questions to put to the prospective partner. They would not establish financial soundness or prove that an undocumented relationship does not exist.

Lawful boundaries and proportionate follow-up

Public visibility is not a blanket permission to collect or use information. Lo Presti excludes unauthorized access, circumvention of access controls and unlawful acquisition of personal data. Privacy and data-protection obligations remain relevant throughout the assignment, including when material can be found through an ordinary search. Collection should stay connected to the agreed requirement rather than expand into unrelated personal information.

Follow-up should respond to a defined gap or a meaningful change. A monitoring assessment could identify which public developments would justify revisiting a conclusion, such as a new filing or regulatory notice. Further research could then test a specific unresolved point within an agreed scope. Where the available sources cannot answer the question, the appropriate outcome is an explicit limitation, not an inference presented as a verified fact.

From search results to a tested account

A useful OSINT brief starts with a precise requirement: which entities, time period, events or relationships need to be understood, and what decision depends on the answer. Clients should expect to be asked what has already been reviewed internally, which languages and jurisdictions are in scope, and whether the output needs to be a one-off briefing or an ongoing monitoring feed, since the two require different collection plans and cost structures.

The analytical process treats provenance as seriously as content: each item is logged with its original source, publication date, and whether it is primary (a filing, a court document, an official statement) or secondary (a summary, aggregation or repost). Apparent corroboration between multiple outlets is checked to confirm they are not all citing the same unverified original report. Entity resolution — confirming that a name, address or registration number refers to the same party across different records — is treated as a discrete analytical step, since conflation is a common cause of error in open-source work.

Deliverables include a chronology with sourced citations, an entity or relationship map showing documented versus inferred links, and an explicit statement of coverage gaps — periods, jurisdictions or topics where public material is thin or unavailable. Constraints are stated rather than smoothed over: some jurisdictions have limited digitised records, some material is available only under subscription terms or access restrictions that must be assessed before use, and the absence of online coverage does not establish the absence of the underlying activity. Automated search tools can surface candidate material, but inclusion in the final report requires human verification of relevance and authenticity.