Stephanie Carvin is an associate professor at Carleton University’s Norman Paterson School of International Affairs, and a former national security analyst with the Canadian government.
With the release of Justice Paul Rouleau’s Public Order Emergency Commission Report into Ottawa’s use of the Emergencies Act, and the serious accusations of foreign interference by China, Canadians are facing the reality of national security challenges affecting their country. And while intelligence documents are playing a crucial role in this reporting, there is not a lot of discussion about what exactly intelligence is, and what its limits are.
Essentially, what we are referring to is reporting that intelligence agencies get from sources, usually human or technical. (Depending on their need, these agencies may also use other forms of information, such as open-source, geo-spatial, even weather.)
Human sources (HUMINT) are recruited because they are deemed likely to have some kind of access to information that helps an investigation. They may assist an investigation out of loyalty to Canada, revenge, a desire for money or some combination of factors. Over time, HUMINT sources are evaluated for both their access to information and the quality of that information. Like all humans, they will sometimes get facts wrong – possibly because they heard a rumour, misinterpreted some detail, or may not know the full picture of a given situation.
Technical sources, such as wiretaps and signals intelligence (SIGINT), provide information obtained through the “global information infrastructure.” This includes emails, texts and other electronic communications. It provides access to conversations that may offer details about a target’s intentions, beliefs, contacts or orders.
When HUMINT and SIGINT information is first obtained, it is “raw”; it has not yet been assessed for its validity or accuracy. It has also not been contextualized alongside other information that an intelligence service may have about a case.
This contextualization is done through intelligence analysis. Desk officers and/or intelligence analysts will look at the different pieces of HUMINT and SIGINT information they have collected and put together a product that provides some kind of summary about an issue or case. Depending on need, demand or the quality of information, the product may be used to inform an investigation at the operational level or brief senior intelligence leaders, other analysts within the intelligence organization or other government departments and agencies, senior policymakers or even allied agencies.
Sometimes these products will be summaries of raw intelligence, without much context, but are deemed important enough to move up the chain. Other times they will be more substantive projects that can take months to put together, with lots of context and assessment.
This context is important for understanding some of the national security issues that are currently in the headlines. When reading reports about leaked intelligence products, we need to be aware of the kind of sources they are based on, or at least acknowledge what we do not know; there is a big difference between “a source said” and “multiple, corroborated sources say.” A source may be experienced, and of known reliability – or they may be new. Agency source evaluations can provide important guidance for interpreting information in these documents, but that may not always be available to journalists.
Imagine you have a friend – let’s call him Jim – who just got back from his trip to Bermuda. Jim enthusiastically tells you that he caught a massive fish in the ocean, won the resort limbo contest and had a wild night with a model on the beach.
Now, you know Jim went to Bermuda. You know he fishes but that he often exaggerates the size of the fish he catches. Moreover, you know Jim has a bad back and probably can’t limbo. And the model? Sure – he met a hand model. So, before sharing Jim’s stories with your friends, you’re going to contextualize the information he’s provided.
Or let’s say you overhear Jim telling all this to another friend. If you were to write an intelligence report about it, you would simply document what Jim said he did. But if you wrote an assessment, you’d put a lot of caveats on that product, and you’d maybe ask his travel buddy to corroborate it.
Understanding the limits of intelligence documents is certainly not to imply recent reports about foreign interference are incorrect. There is clearly a lot of information both in leaked reports and open-source information that suggests there is cause for concern. But understanding its limits does bolster the case for an inquiry into the allegations of Chinese electoral interference, as it would be able to review the totality of intelligence, alongside other information, to better understand the challenge Canada is dealing with.