Reflections on the complexities of modern national security by a data science entrepreneur

Over the last year I have been fortunate to be invited to contribute at several events organised by the Ditchley Foundation. The topics have ranged from international trade, through populism, and in the latest case, a conference on national security and the future of nuclear deterrence. The conference was an impressive gathering of various Ambassadors, national security advisors, and researchers on the topic – and me, a data science entrepreneur with no prior knowledge of national security questions. As such, this summary of my experiences is an outsider’s view on discussions that almost invariably happen behind closed doors, and most of us “regular” citizens know nothing about.

Are we safe?

First questions: do you feel safe, on a national level? If not, why? Is that really true? And should you be more worried? I think at first hand, most of us would answer ‘yes’ on the first question, as a nation-against-nation attack in Western Europe and the US is now relegated to the memories of our parents or grandparents. However, something that was discussed extensively at the conference was the constant, low level attacks that Western democracies endure from hackers and cyber-terrorists on a daily basis. These attacks range from flooding our media with “fake news”, with the intention of influencing the population and ultimately destabilising our democracies, through to outright attempts to harm critical infrastructure. There was a pervasive feeling that the barrage of minor attacks on the foundations of our democracies could, in the long run, lead to greater destabilisation than the threat of nuclear attacks ever could.

A statistical truth

The flood of “fake news” and disinformation is leading to a challenge to identify truths. This, in turn, hampers national intelligence agencies’ abilities to make critical judgement calls on the information they have. Here, I think data science and AI can help. I would urge these agencies, and governments in general, to adopt a “statistical approach to truth”. Whereas agencies in the past had limited, but verified, sources of information, today we have a flood of information, from countless data sources such as satellite data, mobile network data, social media reports, news reports, seismic and geologic data, and many more. Software can be built to create a comprehensive view on, say, an event and based on a threshold of confidence, a truth can be statistically derived.

Manhattan project for AI

In our tech circles, defence and intelligence do not have the best reputation. Many data scientists are motivated by social purposes and prefer not to be involved in anything that may ultimately harm other individuals. The week-end deep dive into national security questions made me realise that the situation is more complex than that. All countries and societies must have a right to defend themselves, when their chosen way of life is under threat. I think here, the wider population has a right for more information on exactly what those threats are, and what measures are being taken to keep us safe. This is particularly important as it is clear to me that governments will need to invest much larger amounts into bolstering AI and cyber defence talent and capabilities. This is an AI arms race, and we need an AI Manhattan project. The only way to motivate the best and brightest tech talent is to ensure that they understand what is at stake. For this, governments need to educate the people.

Are we at risk of anti-globalisation?

One thing that struck me during the conference, and which may be a consequence of the professions of the individuals in the room, was a pervasive feeling of distrust among nations, and a perceived need to make our countries more “resilient” and independent. Stockpiling was brought up in one conversation, for example. As a non-security expert and lover of all things global, I found these conversations worrying. Are we giving up on globalisation? To prepare yourself for conflict, even though swearing that you do not want to act on the preparations, destabilises the conversation, decreases trust, and starts an arms race of conflict preparation. Along with other sentiment in our societies related to populism, nationalism, and protectionism, I fear for the future of globalisation, which would be a great shame for the whole world. We have unfortunately seen the fruits of nationalism too many times before.

The future role of governments and state

Ultimately, will it be the role of the state to keep us safe in the future? The drone attack on the oil refineries of Saudi Arabia on 14 September by, allegedly, a small rebel faction just goes to show that serious damage can now be inflicted on large countries, by non-state actors with limited means. Terrorist organisations can hack the grids of nations and plunge them into darkness with just a couple of hackers with laptops and an internet connection. In the brave new digital world, states have limited control over security. We could turn to private enterprise for help? But for companies like Google, Facebook, or Amazon to somehow come to the rescue, they would need to be convinced there is a threat against their own existence - a hard task. Ultimately, we may just need to accept that the geopolitical landscape is in upheaval, and that the future of nations, governments, and societies will look very different fifty years from now. Until then, we all need to do our best to keep fingers off triggers, and the world safe from nuclear destruction.

My thanks to James Arroyo, Sandra Ricks, and the team of Ditchley Foundation for organising the conference, and for the invitation.