Putting the Steele into intelligence reform

Robert Steele is one of the more interesting writers on intelligence. Based in the US, and a former practitioner he has brought an enormous amount of energy to the questions around intelligence effectiveness and intelligence reform, and can rightly be thought of as a grandfather of the open source intelligence movement, and more recently the expanded “Open Source Everything” meme. I should insert the health warning that he has appeared in the Companion guide that Mike Goodman, Claudia Hillebrand and I edited, so I am not entirely impartial on this, but I would place myself as a ‘critical friend’ of his work.[i]

He has recently published a semi-manifesto piece about US intelligence and it can be found on this link. I have distilled the following key points from it, that I want to write around briefly here, but the original piece is where his take on these issues sit, obviously: 1) intelligence should be about decision support; 2) intelligence is currently being justified along the lines of the quantity of secrets it produces the Executive without regard to the total government need; 3) there is a dominant discourse that only secret intelligence agencies are equipped to ‘do’ intelligence; 4) Parliament and politicians in general desperately need intelligence qua decision-support, sense-making applied to all information secret and open that applies to their functional domains; and 5) the public desperately needs intelligence, again in the form of decision support.  Recently the public has become the object – Americans would say the target – of intelligence agencies, which is quite the opposite of the public being a virtual intelligence network in being, contributing to national and public security more effectively by leveraging the creative commons approach to information, what some call collective or co-intelligence.[ii]

I actually think Steele’s argument needs to be run in reverse chronological order to take on the compelling edge he desires. When I spoke to him a few years ago he was extolling the virtues of connecting up the mobile telephony and tablet computing of the developing world, up to and including equipping them at no cost to themselves, on the grounds that the information they would bring and pool would be invaluable to understanding the politics, security and economics of countries and regions we take best guesses at. He embraced the concepts of, among others, C. K. Prahalad (The Fortune at the Bottom of the Pyramid) and Yochai Benkler (The Wealth of Networks) along with the ideas coming out of Brazil from Puerto Alegre, on participatory budgeting and pedagogies of freedom. The recent and ongoing revelations about mass surveillance (and I would strongly urge anyone reading this to access the European Parliament debates on this – the Guardian published an early synopsis of this work yesterday) obviously opens an alternative reading to the wirelessly wiring up of the developing world. But if we stick with the creative commons notion, it is likely to be the case that we could achieve most intelligence objectives – Steele cites at around 95% – without needing to recourse to covert means. Anyhow, the public has become the object of intelligence agencies, rather than the people they serve. That is a clumsy and sweeping summary, but readers of KoW will know that I think this is becoming the dominant discourse around which the public in particular understand the role of intelligence agencies: their usurping master and virtual enemy rather than a legitimate public service. Readers will also know that I think it would be useful to try and reposition this understanding. I think that Steele is correct that insecurity often flows from a lack of education or understanding at all levels from elite to base – and this in turns comes from failures or fractures in the availability of information and skills and techniques to deal with this information. The meta-analysis he presents here feels right to me – insecurity can be addressed in ways that do not require the secret community, but the devil is always in the detail. Is this actually utopianism, or can something workable be generated?

Steele’s point that the key risks and threats to modern society actually sit in policy areas in which secrecy clearly does not help – eg poverty and health, I think is absolutely correct. Those well worn, and indeed hackneyed debates around ‘securitization’ play exactly into this space – by classifying poverty as ‘security’ do we actually remove it from the sort of intellectual terrain in which we could make progress on it? And by extension, therefore, it might be a useful and efficacious thing to do, to narrow the remit of security to issues and areas that actually benefit from the skills, tools etc of the secret world. Intellectually that does not seem contentious to me, and yet politically – and with the industrial scale of security analysis and expansive tendencies – that is likely to be a controversial and rejected thought.  So, Steele argues that Congress (and for this post Parliament) do need intelligence – decision support – and rather urgently, only now applied to all policy questions across all Cabinet domains. The British Parliament barely receives intelligence (either product nor oversight) as it is, so it’s difficult to see that unless Parliament drags intelligence closer to its ambit that this sentiment applies here, but we might be able to reapply it or reframe it as Parliament would benefit from improving its open source and analytical function (which are already pretty good) and to deal more widely with an open source gathering and analytical function. There are some Parliamentarians for whom this already makes a great deal of sense. That raises the question: if intelligence is decision support, who should do it, at what of confidence, and with whom should the results be shared? Steele argues that secrecy is inherently a decapitation function in conflict with the mission of decision-support, and that the best intelligence is that which can be shared broadly, eg with the media, the public, and all other stakeholders such as the EU.

I think Steele is correct in his first three points: intelligence as decision support, in being justified as quantity of intelligence, and that the craft of intelligence as decision support resides right now, albeit severely impoverished, solely within the secret world. The textbook account of UK intelligence (and this may be a question of scale) is that its function was mostly to provide decision support and not to just warehouse data. Recent technological advances may have slanted that a little, but I think we can hold to that textbook line, and so it might be that we hold a distinction between the US and the UK on this. But I did wonder, having thought about Steele’s piece more, whether the number of policy areas or issues requiring the full machinery of secret government could usefully (and for efficacy reasons) be dramatically reduced to a strong core. An open source fusion centre or similar providing challenge, or working on its own terms might be the first step to making that judgment properly. Indeed, Steele has proposed an Open Source Agency for the USA, a Multinational Decision Support Centre for NATO/EU, and a United Nations Open Source Decision Support Information Network for the world at large.

[i] His chapter, “The Evolving Craft of Intelligence” is free online at his website, by agreement with the publisher.

[ii] Stewart Brand started the meme with the Co-Evolution Quarterly that evolved into the Whole Earth Review. Tom Atlee is the father of the co-intelligence meme, and founded The Co-Intelligence Institute.

This post was originally written in January and is posted with delay because of KoW’s redesign.


2 thoughts on “Putting the Steele into intelligence reform

  1. Returning the focus of intelligence agencies to providing decision support does not in itself mean that the public will be better served-though it may benefit the public finances. Intelligence failures of course are routinely attributed to lapses in analysis not a lack of data. The false promise of open source intelligence is best shown by the great intelligence failure of recent years; the failure of the Federal Reserve Board in 2008 to accurately determine the risks to the American economy. There should be no better example of open source intelligence than the world’s largest and most liquid financial markets, where relevant information is collected by the government and routinely shared with the public yet policymakers as we now know -and was clear from public debates-were divided as to whether inflation or recession posed the greater risk. Groupthink, cognitive closure, interpreting data so that in neatly fits one’s analytic framework -it was all there at the Fed. The information the Fed needed was available yet they failed to provide sufficient warning to Congress and the President as well as to themselves. I am generally shocked by how scholars and commentators on intelligence ignore the literature on financial markets-these people have been thinking about how to get individual actors to reveal private information and the ability of analysts to predict trends for a long time. Financial market behaviour tells us that individual actors are always looking for an information advantage-the key insight that isn’t available to the market participants, and they will pay a hell of a lot for it. This doesn’t mean that it’s worth price and government should examine whether there can be cost savings by relying on open sources. Logically it follows that intelligence will be shared across government at the discretion of the agency paying for its production. Again there is public choice literature on budget maximizing bureaucracies. The emphasis on intelligence collection obviously arises from a demand for information but more importantly the supply of new tools that, given an accommodating legal framework, promise to answer hitherto unknowns, increasing the confidence in analytic judgement. Scaling down the size of the intelligence establishment does not necessarily mean that the balance between analysis/decision support vs. collection will be altered.

  2. Vladimir,

    I agree with all that you say but would add an important distinction: intelligence without integrity is not intelligence. The OSINT movement was hijacked by lazy government bureaucrats and corporations focused on optimizing overhead fees but telling the government that only citizens with clearances, “butts in seats,” were the solution. In fact the opposite is true. OSINT done properly first gets the question right, then goes out and skims the cream from who knows best in the moment. I salute Stefan Dedijer, who at OSS ’92 shouted out, “KNOW WHO KNOWS.”

    I wrote a preface for my quick book, free online, ELECTION 2008: Lipstick on the Pig (2008) entitled “Paradigms of Failure” in which I made the point that all eight tribes of intelligence gave up their integrity: academia, civil society (including labor and religion), commerce, government, law enforcement, media, military, and non-government/non-profit.

    One reason I morphed from being a champion of OSINT to a champion of Open Source Everything (OSE) is because I realized that integrity only flourishes in open systems where all minds can share information and share the burden of sense-making. All of the industrial era systems are closed systems whose leaders are focused on optimizing their personal profits and privileges, NOT on the public interest. OSE is affordable (largely free), inter-operable, and scalable, which cannot be said of any of the propriatery solutiosn that seek to “capture” clients at any cost and regardless of the opportunity cost.

    I am actively seeking invitations to speak out on these matters. My new bio with links at the bottom and international statements of support can be found at http://tinyurl.com/Steele2Speak

    With best wishes,

Be sensible, be polite

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>