We could soon face a robot crimewave

[Kevin Lancaster]

We could soon face a robot crimewave

This is where we are at in 2017: sophisticated algorithms are both predicting and helping to solve crimes committed by humans; predicting the outcome of court cases and human rights trials; and helping to do the work done by lawyers in those cases. By 2040, there is even a suggestion that sophisticated robots will be committing a good chunk of all the crime in the world. Just ask the toddler who was run over by a security robot at a California mall last year.

 

[OneKnight]

It is going to be a while before this is going to be a concern though.

[video=youtube;fpF-3orhnjg]https://www.youtube.com/watch?v=fpF-3orhnjg[/video]

 

[Arthur]

For classical Aristotelean Thomists such as myself, robots are radically doli incapax, incapable of malice, deceit and thus guilt.

In other words, there can be no such thing as a robot crime in the classical and traditonal legal or moral sense of the term.

This is because for an external action (what lawyers call an actus reus) to be a crime also requires men's rea or intentionality.

Philosophically speaking, intentionality is the product of two distinct but not separate faculties or powers, namely intellection and will. So, culpbability requires both knowing (intellect) and choosing (will). Culpa is essentially a defect of the intellect, whereas dolus is described by jurists as a defect of the heart, which rests in the will rather than the intellect per se.

For the sake of the argument one can grant that some future robot AI 'brain' might be functionally indistinguishable from a human intellect, or even appear to be superior to it, but it is radically incapable of a will, let alone a free one.

In other words, an AI robot might have a locus for a machine intellect (scripts, algorithms database, etc), and thus be culpa capax. But there is no locus for a machine will, hence it is doli incapax and incapable of criminal liability.

Therefore it cannot ever develop a free choice, the abuse of which is at the heart of criminal capacity. It's why we don't put animals on trial and hold them accountable for their actions, no matter how smart or 'trained' they are.

When an AI robot commits an actus reus that for humans would entail criminal liability, we will always look beyond the machine to find the human will - either criminal or negligent - that resulted in the machine acting as it did.