Who Owns the Creation of an Artificial Intelligence?

Article Placeholder Image
By Jonathan R. Tung, Esq. on August 22, 2016 11:57 AM

This question is becoming increasingly relevant every day: who owns the product of an artificial intelligence?

Why, the owner of the machine, of course. But is that answer really quite so obvious? After all, who owns the machine if the machine itself is difficult to define? And even more curious, can an intelligence be owned? And should it?

AI Is Here

Quasi-versions of artificial intelligence are already around us. We're on the cusp of self-driving cars, self-directed surgery machines, and machines that we can literally have a conversation with. And of course, as with any new technology, a few kinks have to get ironed out.

Artificially "intelligent" machines already are used to to do quite a bit, including creating algorithms to solve problems. This form of "learning" is making machines more "intelligent" every day. Just ask Watson.

In a few years, we'll be having such a deep and intimate discourse with our machines that we will be spending more time with them than we are with other human beings. Oh, wait -- we already do.

Legal Questions

Current intellectual property law in this country does not allow and does not recognize machines as "individuals." The legal fiction has not yet been invented. Fine, so machines don't own what they make.

That basically means the question of ownership of made product is really between other parties and those who created the machine who created the product. But if one looks to the language "discovered" or "invented," they'll discover that a fair bit of disagreement exists.

Current Laws Not Equipped to Handle the Question

For example, does someone discover the AI, when he or she single-handedly develops it on his or her PC? What about those situations when multiple people contribute to the creation of the intelligence? And things get even worse legally when you consider licensing and persons who "train" intelligences to move beyond their original specifications.

But if human "inventors" are to take ownership of the objects created by their creations, they ought to also bear the responsibility of the artificial intelligence's violation of other's patents. This would mean that any time a machine makes a new product that is in potential violation of an existing patent, the humans that contributed to its existence could be on the hook.

What About the Ethics

If machines become more intelligent and humans own the machines, won't the humans essentially become slave-owners? This sort of issue was on the minds of some academics decades ago. Actually, it's not unfair to say that we'll have to grapple with the legal and ethical questions very soon -- and, as usual, we're ill equipped to tackled it.

After all, technology is being worked on now that will give machines the ability to read emotions -- a uniquely organic intelligence capability.

Related Resources: