THE 6P METHOD FOR COMPREHENSIVE INFORMATION PROCESSING AND VERIFICATION Author: Larysa M. Gavriley

• 

The name “6P” is retained in English as the author’s established designation. It is the Latin-script rendering of the original Ukrainian name “6П”, in which the names of all six stages begin with the letter П. The stage codes P1-P6 remain unchanged, while their English names are translated according to their meaning.

The 6P Method is an author-developed six-stage framework for tracing, verifying and documenting information, and for assigning responsibility before the information is used in a professional conclusion or decision.

The method was developed in response to changes in the information environment associated with the large-scale use of generative artificial intelligence. Digital materials can now be copied, translated, abridged, recombined and automatically rewritten at high speed. As a result, a statement may retain a convincing form while losing a clear connection to its origin, primary evidence and original context.

The method treats every item under review as an information object. An information object may be a document, dataset, table, image, audio or video recording, database entry, public statement or AI-generated answer. The main unit of review is a decision-relevant claim: a factual, legal, scientific or other assertion on which a conclusion or decision depends.

The method consists of six consecutive stages:

P1 — Provenance

Control question: Where did this information come from?

This stage establishes the author or source, creation date, initial publication, method of creation, source materials, involvement of automated systems or artificial intelligence, and subsequent changes or versions. The result is a complete, partial or unknown chain of provenance.

P2 — Primary Source

Control question: What exactly supports the claim?

Each decision-relevant claim is linked to its primary source and to the precise passage, claim, paragraph, figure or other location that provides the evidence. A reference is treated as a route to evidence, not as evidence in itself.

P3 — Verification

Control question: Does the cited evidence actually support the claim?

This stage tests the correspondence between the claim and the evidence. The depth of review is proportionate to the likely consequences of an error. Facts must be distinguished from interpretations and professional conclusions.

P4 — Disclosure of Transformations

Control question: How was the material created or modified?

This stage discloses material transformations and limitations, including the use of artificial intelligence, automated editing, translation, abridgement, compilation, incomplete data and limits of verification.

P5 — Protocol

Control question: Can the verification process be reproduced?

The protocol records who worked with the material, which claims were checked, which sources were used, how the verification was performed and which version was approved. Another competent person should be able to understand the process and, where practicable, reproduce it.

P6 — Personal and Institutional Responsibility

Control question: Who is accountable for the result?

This stage identifies the competent person or institution that checked the material, approved it, authorised its use and is responsible for correcting an error.

The central rule of the 6P Method is that no stage can replace another. Source identification does not replace verification. Disclosure of AI use does not establish accuracy. A complete protocol does not make an unsupported conclusion reliable. Responsibility cannot replace evidence.

A serious failure at any stage is treated as a critical-link failure. Information cannot be classified as substantiated merely because the remaining stages were completed.

The method uses a risk-based approach. Routine language editing may require review by the author. Public, commercial and analytical materials require verification of names, dates, figures, quotations, source identity and currentness. High-consequence uses, including law, medicine, finance and security, require competent professional review, primary sources, documented verification and, where practicable, independent confirmation.

The final result is classified as:

Substantiated information — critical claims have a sufficient evidentiary basis, the required review has been completed, material transformations have been disclosed and responsibility has been assigned.

Information with stated limitations — a material uncertainty or gap remains, but it has been identified, recorded and reflected in the conclusion.

Unsubstantiated information — a critical claim lacks sufficient evidence, conflicts with the primary source or has not undergone the required review.

An unsubstantiated result does not necessarily mean that the claim is false. It means that the available evidence is insufficient for the claim to be used as an established fact.

Within the 6P Method, an AI-generated answer is not a primary source. It is an information object that must be checked according to the purpose and risk of its intended use. The relevant question is not whether artificial intelligence can be trusted in the abstract, but what verification a particular output requires before it is relied upon in a particular professional context.

APPLICATION TO PATENT SEARCH

In patent searching, the 6P Method operates as a separate evidentiary control layer. It does not replace the legal criteria for patentability and does not alter the meaning of the X, Y and A relevance categories.

X/Y/A classification determines the patent relevance of a document. The 6P Method determines whether that classification and the resulting conclusion are supported by a traceable and reproducible evidentiary chain.

The integrated sequence is:

Search → X/Y/A classification → 6P control → professional conclusion.

For every key document, the searcher should identify the exact claim, paragraph, figure or other passage on which the conclusion depends. Titles, abstracts, semantic similarity, machine-generated summaries and automated rankings may guide the search, but they do not replace examination of the underlying disclosure.

Machine translation, optical character recognition, automated summarisation and other transformations must be identified where they may affect meaning. The protocol should state which databases, classification systems, languages, queries, dates and search routes were used, and where the limits of the search remain.

A negative result must be confined to the search that was actually performed: “not found” does not mean “does not exist”.

Artificial intelligence can expand terminology, suggest search directions, group documents and accelerate preliminary review. A competent professional remains responsible for the relevant date, the technical meaning of the disclosure, the classification of the document, the sufficiency of the evidence and the limits of the final conclusion.