Evaluating Information Sources

Every intelligence assessment is only as reliable as the information on which it is based.

No amount of analytical skill can compensate for poor-quality information. If inaccurate, misleading or unverified information forms the foundation of an assessment, the conclusions drawn are unlikely to be reliable, regardless of how sophisticated the analysis appears.

This is why evaluating information is one of the core responsibilities of an intelligence analyst. Before identifying patterns, drawing conclusions or making recommendations, analysts must first ask a much simpler question: How much confidence should I place in this information?

The answer is rarely straightforward.

Information comes from a wide variety of sources, each with its own strengths and weaknesses. Some have consistently demonstrated their reliability over many years, while others may have no established track record at all. Even highly reliable sources occasionally report inaccurate information, just as an unknown source may sometimes provide entirely accurate intelligence. The analyst's role is not to accept or reject information outright, but to evaluate both the source and the information before determining how much weight should be given to it during analysis.

Why intelligence is graded

Evaluating information is only part of the process. Intelligence also needs to be communicated in a way that allows other people to understand how reliable it is. This is the purpose of intelligence grading.

Intelligence grading provides a common language that allows investigators, analysts and decision-makers to quickly understand how much confidence should be placed in the underlying information. Importantly, intelligence grading does not determine whether information is true or false. Instead, it represents an assessment of both the reliability of the source and the credibility of the information at the time it was recorded.

These are two separate considerations. A source may have a long history of providing reliable reporting but still make mistakes, misunderstand events or unknowingly repeat inaccurate information. Equally, a source with no previous reporting history may provide completely accurate information on a particular occasion. Professional analysts therefore assess both the source and the information independently before deciding how much weight should be given to the report.

One practical benefit of intelligence grading is that it supports the safe sharing of information. Before intelligence is disseminated, it is often sanitised by removing details that could identify a source or reveal sensitive investigative methods. For example, rather than stating that Betty at number 9 witnessed an assault, a sanitised report may simply state that a witness reported the assault. While this protects the source, it also removes valuable context. Intelligence grading preserves that context by communicating the analyst's assessment without revealing the source itself. Sanisatisation also protects investigative capabilities as it masks whether the information came from a trained surveillance officer, an anonymous member of the public or an untested confidential source.

Standardised grading also creates a shared language between organisations. Without a common grading framework, there is a risk that one organisation interprets information very differently from another. Whether intelligence is exchanged between police forces, government departments or private-sector investigators, the recipient can understand the originating analyst's assessment without needing to know the full circumstances surrounding the collection.

Corroborating information

Few pieces of intelligence should be accepted in isolation. Confidence increases when information can be corroborated through independent sources, but analysts must be careful not to mistake repetition for corroboration.

This is particularly important in OSINT investigations. During breaking news events, dozens of websites may publish identical reports within minutes. At first glance, this appears to represent multiple independent sources. In reality, every article may trace back to a single anonymous social media post. The number of reports is far less important than the number of genuinely independent sources.

Analysts should also consider whether information originates from a primary or secondary source. Primary sources provide first-hand information by directly witnessing an event, creating a document or recording a transaction. Secondary sources report information obtained from someone else. Each additional layer between the original event and the analyst introduces opportunities for misunderstanding, omission or distortion, making access to the original source preferable wherever possible.

Formal intelligence grading systems

Several formal grading systems exist to help communicate the quality of intelligence, which tend to separately assess source reliability and information credibility.

Internationally, one of the best known is the Admiralty Code, which separately evaluates source reliability and information credibility. Within UK policing, intelligence has historically been graded using the 5x5x5 model before moving to the simplified 3x5x2 system under the National Intelligence Model.

Under the Admiralty Codesource reliability is commonly graded using letters ranging from A to F, while information credibility is assessed numerically from 1 to 6.

An A1 report would represent information from a completely reliable source that has been independently confirmed. At the opposite end of the scale, an F6 report would indicate information from an untested or unreliable source that cannot currently be assessed.

Although these systems differ in presentation, they all share the same objective: providing a consistent and auditable assessment of the quality of intelligence. The specific grading system matters less than the principle behind it, which is clearly distinguishing the reliability of the source from the credibility of the information.

Evaluating publicly available information

The growth of OSINT has transformed intelligence collection, but it has also challenged some traditional approaches to intelligence grading.

Historically, some organisations adopted the practice of assigning the same intelligence grade to almost everything collected from the internet. The reasoning was understandable. Online information can be anonymous, manipulated and difficult to verify.

However, treating every internet source as equally reliable ignores the enormous variation in the quality of information available online. A Companies House filing, a government publication, a verified corporate website, an academic journal, an anonymous forum post and an AI-generated social media account are all internet sources, yet clearly they should not receive the same assessment.

The internet is simply the medium through which the information was accessed. It should not determine the intelligence grade by itself. Instead, analysts should evaluate online information in exactly the same way they evaluate any other source by considering who produced it, how it was obtained, whether it can be corroborated and whether there are reasons to doubt its authenticity.

Absence of evidence is not evidence of absence

Another common mistake is assuming that because no supporting information has been found, something did not happen. In intelligence work, missing information often reflects collection limitations rather than objective reality.

A lack of social media activity does not necessarily mean an individual was not present at an event. Likewise, the absence of criminal records does not demonstrate that criminal activity has never occurred.

Analysts should therefore distinguish carefully between no evidence and evidence of nothing.

Confidence is not the same as grading

One area that is often overlooked is the distinction between intelligence grading and analytical confidence.

Intelligence grading evaluates the source and the information available at the time it was collected. Analytical confidence reflects how strongly the analyst believes the final assessment is supported by the totality of the available evidence.

For example, a verified social media account may provide reliable information that was genuinely published by the account owner. However, an analyst may still have only moderate confidence that the account actually belongs to the individual under investigation until it has been corroborated through additional evidence.

Similarly, two pieces of intelligence may receive identical grades but contribute very differently to the analyst's overall assessment depending on how they fit alongside other available information.

Recognising this distinction encourages analysts to think beyond individual intelligence reports and consider the strength of the overall intelligence picture.

A Practical Framework for Evaluating Publicly Available Information

Professional analysts rarely evaluate information using a single question such as "Is this source reliable?" Instead, they consider several characteristics of both the source and the information before deciding how much weight it should carry within an assessment.

A useful framework is to consider six questions:

1. Is it complete? Does the source provide sufficient context to evaluate the information? Can you identify the author, organisation, publication date and purpose?
2. Is it credible? Does the source have the expertise or authority to make the claims presented? Has it demonstrated reliability previously?
3. Can it be corroborated? Are the claims independently supported by other reliable sources, or are multiple reports simply repeating the same original information?
4. Is it relevant? Even accurate information has little value if it does not contribute towards answering the intelligence requirement.
5. What impact does it have? Does this information support or challenge your current assessment? Does it suggest a new hypothesis or identify an intelligence gap?
6. Is it accurate? Is the information current, factually correct and internally consistent? Has anything changed since it was published?

Good intelligence begins with good information

Analytical techniques, visualisations and well-written reports all add value, but they cannot compensate for poor-quality information. Every assessment should begin by evaluating the reliability of the source, the credibility of the information and the confidence that can reasonably be placed in the conclusions drawn from it.

The best analysts are not those who accept information most readily, nor those who dismiss everything with scepticism. They are those who approach every new piece of information with disciplined curiosity, continually asking where it came from, how much trust it deserves and what additional evidence is needed before reaching a conclusion.

Only by building intelligence on a solid evidential foundation can analysts produce assessments that genuinely support informed decision-making.