AI-controlled espionage – a Bond film or reality?

However, it is undeniable that recent technological revolutions – especially that Internet and the artificial intelligence (AI) – make our world more and more what was called “cyberpunk” in the 80s.

What some might find an unexpected twist, but I’d rather call “prediction from the past,” is that the world’s intelligence agencies have begun actively using artificial intelligence to wage their (now virtual) spy wars.

And now we live in a world where John Badham’s film WarGames less like a fiction and more like a documentary (albeit in color).

Warfare in cyberspace

We can all see the effects of the Cold War shifting between world powers to virtual space.

In recent years, the number of cyber attacks threatening or even damaging the largest government agencies has increased to unprecedented levels. According to a source this value increased by 95% in the last half of 2022.

The more we network, the less protected we are by what traditionally gives security to every nation: military power and security.

Maybe it’s rival spy agencies fighting each other, but the consequences are pretty dire.

Criminals can now reach any border without firing a single bullet and without having to cross a mountain, river, desert or border fence.

And you no longer have to have the full might of the Chinese or Russian army to pose a serious threat to a US government agency.

What you need is a small group of experienced people hacker, equipped with the correct virtual weapons and possibly assisted by AI. In short: War is not what it used to be – especially not a war of espionage.

AI as a weapon in espionage wars

As the weapons used in espionage wars have evolved, so has the nature of the war itself.

Intelligence today has an entirely different shape, size and scope as millions of data points are being collected at every moment.

The technologies for big data likely to revolutionize all espionage scenarios, especially as most of this data is open source and can be collected by anyone rather than a few secret agents.

Information travels much faster, and the number of stakeholders collecting that information has increased significantly.

Facebook, Twitter and Google process hundreds of petabytes of data every day. At the same time, new commercial surveillance satellites are regularly launched.

Everything is much faster, and the response times are significantly shorter.

Amy Zegart, Chair of the HAI Steering Committee for International Security, commented: “During the 1962 Cuban Missile Crisis, President Kennedy had 13 days to secretly consider what he would do after U-2 spy planes spotted Soviet missiles in Cuba.”

“On September 11, President George W. Bush had only 13 hours to review information about who was responsible for this horrific attack and how the US should respond.”

“Today the decision time could be 13 minutes or less.”

With so much data to process and so little time to do it, AI is more of a necessary tool than a luxury.

Some of the biggest global players have already understood this and with the collection of records started to feed their AI – with or without the consent of the recipients.

This is how MI6 chief Richard Moore explained his suspicion that China set up “data traps” during the recent Covid-19 pandemic: At that time, Beijing is said to have often forced the countries that bought vaccines from China to make their vaccination data available.

According to Moore, these moves should ring alarm bells about Chinese willingness to “use coercion or create dependencies” that pose a threat to the international order.

Furthermore, Moore explained how the British secret services are currently using AI to cut off the flow of arms to Russia in the current war against Ukraine.

Of course, being on the “good side” of the espionage war and all good spies, they will only use AI in the “most ethical and safest” ways. Yes, of course.

And now we must deal with the question. Which question? The question that is always asked when it comes to AI:

Will human spies be replaced by AI?

I’ll admit I’m a bit biased as I’m a data analyst myself. But the world of espionage is broader than my field of work, so I can try to give a hopefully neutral answer to this question.

According to figures such as Moore and Zegart, there is no such risk because in many ways the human factor cannot be replaced.

It is true that the collection of information will increasingly depend on the use of AI.

The amount of Datathat needs to be processed is just too big for humans alone. This is especially true if you only look at the open source data.

As in many other industries, AI will be another tool to facilitate and speed up the sifting through of these vast amounts of information.

Humans are still required to make sense of this data and turn it into actionable “intelligence”.

For example, AI could be used to analyze satellite imagery and detect anomalies, saving people the tedious task of examining thousands of images every day.

Even if AI can detect an unusually high level of traffic in an enemy base, a human must understand the significance of this flow of vehicles and decide what action (if any) needs to be taken.

While it is true that data from open sources can be collected more efficiently by AI, the same cannot be said for sensitive information obtained in unconventional ways.

For this type of information retrieval, the human relationships established by agents in the right places can never be replaced by AI.

Conclusion

Using this technology in espionage can greatly improve the capabilities of human spies and intelligence agencies without weaponizing AI.

Therefore, it will be (and to some extent already is) crucial to protect vital national assets, sensitive secrets and technology, and to ensure global peace and stability.

Oh, and Matthew Broderick was in WarGames great by the way.

Crypto exchanges with the lowest fees 2023