AI Capone: The Criminal AI


Yes you got it…. it wasn’t Al Capone, it was AI (Artificial Intelligence) Capone – although what I am about to share with you should concern you regardless.

You probably don’t realize but have you noticed how technology has added a new level of crime related stress in your life. Do you worry about getting hacked – your business secrets and personal life exposed to the world? Do you get concerned every time you use your credit card online? Do you check your bank balance frequently and monitor your credit card bills closely to make sure that you are not a victim of cybercrime? Most likely you have deployed firewalls and other security software and you make sure they don’t expire? Have you ever lost your computer files to a malicious virus?

I know my answer to all of the above questions is “yes” and it is likely that your answer to the above questions will also be a “yes”. Doesn’t this add so much stress for us?

After all … we live in a pretty vicious world. Cybercrime is so dangerous and malvolent that it has added a tremendous burden on us (psychological, financial, and physical).

A few years back I helped develop the strategy and a big data centric approach to secure the world’s largest financial services company and that project truly opened my eyes to the world of cyber security. I was shocked to realize the lethal potential of these crimes. Put it this way… it can destroy lives, change political outcomes, crash economies, kill patients, and casue major accidents.

And that was all before artificial intelligence based agents enter the picture.

Now imagine all of the above crimes – but their power compounded with the non-human, intelligent, autonomous or semi-autonomous, self-directed or human-guided, virtual and physical agents.

For simplicity sake, let us remove the arbitrary distinction between physical (robots) and cyber (software). I am sure the first thing you noticed is that crime with artificial intelligence has two components that are different – first, the obvious one, is that we will be facing intelligent, and potentially self-guided and self-learning, agents – and second, that unlike cybercrime which manifests in virtual setting (even though its impact can be both in virtual or physical setting) the AI based crime will manifest in a host of settings including computers (cyberspace), drones, robots etc. This means both the variety and intensity of crimes can rise significantly.

No wonder Tracey Follows, chief strategy and innovation officer at The Future Laboratory forecasted that by 2040 more crime will be committed by machines than by humans. To read the story click here.

The important thing to consider here is to keep an eye on how legislation, regulation, and liability scenarios develop for the providers of AI software and services. More about that in future posts.