Legal Questions About AI

I’ve blogged about the Law and the Multiverse site before, but I have to link to them again. Today’s post is about Non-Human Intelligences and there’s a whole shelf full of potential SF novels just in this one paragraph:

If the program is a person, is powering down the computer on which it is running murder? Does a powered-down AI have a right to be powered back on? If the program is copied, do we now have two people? Is deleting one of the copies homicide? If the program is installed on a person’s computer against their will, do they have to take care of it forever or can they delete it? Does introducing a trojan horse constitute assault, or is trespass a better analogy? Is the essence of the being the code, the running program, or what? Does the consciousness reside on the hard drive or in RAM? Maybe the CPU cache? What if the program is installed on a really slow computer? In a theoretical sense it’s still the same program, but is it still intelligent? We don’t even really think about regular programs with this kind of rigor, so something as sophisticated as an AI is likely to make a court draw a bright line that means it doesn’t have to think about that stuff.

This is a great site, and should be on the RSS feed reader of everyone who writes SF.


Published by


Angela Benedetti lives in Seattle with her husband and a few thousand books. She loves romance for the happy endings, for the affirmation that everyone who's willing to fight for love deserves to get it and be happy with someone. She's best known for her Sentinel series of novels, the most recent of which is Captive Magic.