A good piece by @gordoncorera but which hits some of my trigger points on how national security is covered. https://www.bbc.co.uk/news/uk-52122991

First, for some discussion on broad definitions of national security, you can do worse than David Omand's 'Securing the State'. It talks about popular perceptions of personal security, though we need to have boundaries at some point, otherwise everything becomes 'securitised'.

The article rightly differentiates between covert collection and broader analysis of populations. I am sceptical that the latter is an 'intelligence' task, at least at the moment.

First, because intelligence collection is a focussed activity. At its most useful it covers individuals, units and organisations. We don't - yet - have the ability to collect intelligence on whole populations and predict outcomes.

Hence why predicting elections is not really an intelligence task, even though analysts are always asked the question. We have a fairly mediocre record in predicting Western elections or referendum results when we have masses of data in our own language!

Second, because if you are talking about openly available information, using a covert collection capability is a waste. The UK community has to ruthlessly prioritise targets. Anything that can be openly collected is a diversion of covert capabilities.

This brings up the other major point. Intelligence is more than 'spying' (espionage). For challenges like these the burden falls upon *analytical* organisations which rarely get a mention in national security stories.

The responsibility for making sense of both open and covert information falls to organisations like the Joint Intelligence Organisation, Defence Intelligence or the Joint Terrorism Analysis Centre.

So the implication of the article is about whether these organisations can, or should, be doing more of this kind of work. They are the ones where the impact of artificial intelligence might be most profound. And then we will need a discussion on whether this is 'security' work.

There are many parts of government which might (or do) baulk at, for example, Defence Intelligence, doing population analysis. And there are scientific areas of excellence which will be better suited than political-military analysts. Security organisations can also over-classify.

So this could be an opportunity for 'Fusion Doctrine' and those in cross-government analytical centres, for example GO Science, to take a look at how they work and share information, taking a broader an less prescriptive view of 'security' issues.

Side note: the CIA crops up a lot in these conversations because it is not just an espionage/covert action organisation. It has a formal role in all-source analysis. SIS in the UK does not, and its relationship with the CIA is closest in the fields of operations and humint.

Extra pedantic side note: yes, it still bugs me that people write 'MI6' when it is SIS and has been for nearly 100 years. I don't care 'that's how people know it', it isn't correct and we shouldn't pander to popular ignorance.