Robots behaving badly: legal responsibility in the age of AI
1. R O B O T S B E H A V I N G B A D L Y : L E G A L
R E S P O N S I B I L I T Y I N T H E A G E O F
A R T I F I C I A L I N T E L L I G E N C E
D R A N D R E S G U A D A M U Z , U N I V E R S I T Y O F S U S S E X
@ T E C H N O L L A M A
2. I , F O R O N E , W E L C O M E
O U R N E W R O B O T I C
O V E R L O R D S
D R A N D R E S G U A D A M U Z , U N I V E R S I T Y O F S U S S E X
@ T E C H N O L L A M A
4. … A N D A
S T O R Y …
• Rule 34 by Charles Stross
(sequel to Halting State).
• Rule 34 refers to an
Internet meme that posits
that there is Internet
pornography for every
conceivable subject.
• Book describes a rogue AI
that starts killing people
with IoT devices.
5. T H I N G S I W I L L N O T
C O V E R
A L G O R I T H M S
6. T H I N G S I W I L L N O T
C O V E R
R E G U L A T I N G S E L F - D R I V I N G C A R S
7. T H I N G S I W I L L N O T
C O V E R
T R O L L E Y P R O B L E M
8. T H I N G S I W I L L N O T
C O V E R
3 L A W S O F R O B O T I C S
9. T H I N G S I W I L L N O T
C O V E R
K I L L E R D R O N E S
10. D O U G L A S H O F F S T A T E R , G Ö D E L , E S C H E R , B A C H
“AI is whatever hasn't been done yet."
13. D E F I N I N G
A R T I F I C I A L
I N T E L L I G E N C E
• “AI is the study of agents
that receive precepts from
the environment and
perform actions.” Russel
and Norvig (2016)
• Agency is the key operator,
so in this talk I prefer to talk
about autonomous agents,
as the “smartness” of the
system becomes
secondary.
15. S E L F - D R I V I N G C A R S
U B E R C R A S H
16. S M A R T C O N T R A C T S
P A R I T Y W A L L E T
17. M A C H I N E L E A R N I N G
D A T A M I N I N G F O R M U S I C
18. S H O P P I N G B O T
! M E D I E N G R U P P E B I T N I K
19. L E G A L
R E S P O N S E S T O
N E W
T E C H N O L O G I E S
• Ban
• Regulate
• Self-regulate (or do
nothing)
• Co-regulate
• Apply existing law
• Draft new legislation
20. S E L F -
D R I V I N G
C A R S
• I don’t think this poses a lot of
legal challenges in the liability
side, 3 types of potential liability
• Product: Manufacturer liable for
making a defective product.
• Service: software developer,
ISP, repair company liable for
causing a fault in the system.
• User: liability through misuse of
the device, or neglect in keeping
it in order (missing a vital patch).
21. S M A R T
C O N T R A C T S
• Similar analysis, product,
service, user.
• Liability made more difficult
due to distributed and
decentralised nature.
• Who caused the bug? Can
they be identified? Is there
recourse? Can you sue
someone to get your
money back?
22. D A T A M I N I N G
• Different type of liability.
• In copyright, what matters is that
someone has used a substantial
part of a work, and that there
can be a connection between
the original work and the alleged
infringement.
• In few systems (US fair use),
most derivative works would be
permitted as transformative, the
same is not true in other
jurisdictions.
23. D O W E N E E D
A N E W
S Y S T E M
• Existing laws on negligence still fit for
purpose for most cases (Tort, delict,
extra-contractual liability).
• Proximate cause in tort of negligence,
an event sufficiently related to an injury
that the courts deem the event to be the
cause of that injury.
• Criminal liability is still problematic, but
unlikely as there is no mens rea.
• We may need an overhaul in contractual
law arising from autonomous agents.
• Revisit the Roman law of slavery?
Servus non habet personam.
25. @ T E C H N O L L A M A
“We seem to be made to suffer. It's our lot in life”
C3P0
Editor's Notes
I always start my presentations with an apology, and this is not an exception. I have presented at re:publica for the last couple of years, and usually I try to do so with a paper in which I am starting to do some research. This is no exception, I have just started being interested in the liability side of AI, instead of the rights given to creations generated with AI. I’m not a technologist so this is mostly my legal mind looking at the field. I am also not going to give any solutions, this is not a regulatory paper, it’s mostly a description of existing laws.
Algorithms are dealt with in detail by many other people, so I will not be covering in any detail here. I highly recommend the work of my colleagues and friends Michael Veale and Lilian Edwards in this area.
The actual regulation of self-serving cars does not interest me, my guess is that this will happen and regulatory efforts will be secondary.
That also goes for the trolley problem, which is an interesting thought experiment, but not what I believe will be relevant.
And the 3 laws of robotics are also a useful science fiction experiment, but not something that I find useful.
And the least I talk about killer drones and autonomous weapons, the better. I think this is a subject for other areas of law, including the law of conflict and
One of the issues we are having is that we tend to frame the debate on the basis of AI as this incredibly powerful tool, almost super-human in speed and capabilities.
The reality however is more mundane. So what interests me is perhaps not so much the killer robot kind, or the evil AI who will enslave humanity feared by so many, but the more mundane kind, which will be the most pervasive. What are the legal implications of a Roomba?
Cyclist Elaine Herzberg was struck by an Uber self-driving in Arizona in March 2018. Police said “It’s very clear it would have been difficult to avoid this collision in any kind of mode (autonomous or human-driven) based on how she came from the shadows right into the roadway”. Latest report is that Uber will be settling the case for a huge amount of money, and the family of the victim will not sue.
On November 6 2017, a bug in a multi-signature smart contract for an Ethereum-based wallet resulted in the freezing of $300 million USD in ETH at the time. Computer says no.
Right now we have large number of companies training their AI using music, text, poetry, etc to produce new works. To give you an example, a program called “Bot Dylan” generates music after being “trained” by listening to thousands of Irish folk songs. Are these resulting works infringing copyright?
In October 2014 the Darknet Shopping bot, an art project that purchased random items from Dark Web markets, purchased 10 yellow ecstasy pills.
There are things that we know we don’t know, and there are things we don’t know we don’t know. We need to see how technologies are applied in practice, as things stand, I contend that most liability arising from autonomous agents can be