Communications of the ACM

systems in combat makes it difficult to apportion blame when something goes wrong. The question of whether the weapon designer, deployer, or indeed, any other entity should take the blame is far from answered. Schwitzgebel suggests the diffusion of blame could be consistent with governments collaborating on undesirable uses of autonomous systems.

Lack of predictability could also
become a more serious threat as AI
systems become more complex. “Hu-
man soldiers in warfare can be unpre-
dictable, but within limits as military
commanders have an understanding
of what has happened in various con-
ditions in the past,” says Schwitzgebel.
“Autonomous systems could be more
unpredictable than humans, which in
warfare could lead to disastrous con-
sequences. The ethics of autonomous
weapons and issues of AI and philoso-
phy are not as widely talked about as
they should be.”
With many questions about the
benefits and dangers of LAWS still up
in the air, and no international agree-
ments in place to provide answers, the
Campaign to Stop Killer Robots and
other research organizations keen to
ensure a ban on their development,
manufacture, and deployment, are
exerting pressure on governments to
adopt and implement their approach.

While the campaign is concerned that the window of time to reach agreement on a ban on LAWS is closing as autonomous weapons are being developed, progress in its favor is beginning to be made and autonomous weapons are moving up the agenda following a December 2016 United Nations review of the CCCW.

Making a small piece of history, the U.N. voted during the review to start a formal process that might lead to a ban on LAWs. Of course, there are no guarantees that the process will be successful, but as Walsh puts it: “If states hadn’t voted to start the process, there would have been no chance to finish.” Russia abstained from the vote.

Countries participating in the
vote agreed to set up an open-ended
Group of Governmental Experts that
will discuss nations’ concerns about
LAWS and the line between auto-
nomous and non-autonomous weap-
ons. The group will meet for two
weeks in August this year, but the ex-
pectation is that it will take multiple
years to reach consensus and add a
protocol to the CCCW that will ban
autonomous weapons operating be-
yond the boundaries of international
humanitarian law.

Further Reading

Losing Humanity: The Case against Killer Robots November 2012, Human Rights Watch and International Human Rights Clinic at Harvard Law School https://www.hrw.org/report/2012/11/19/ losing-humanity/case-against-killer-robots Views of the International Committee of the Red Cross (ICRC) on autonomous weapon systems April 2016, ICRC https://www.icrc.org/en/document/views-icrc-autonomous-weapon-system Three in Ten Americans Support Using Autonomous Weapons February 2017, Ipsos http://www.ipsos-na.com/news-polls/ pressrelease.aspx?id=7555

IEEE Ethically Aligned Design Document Elevates the Importance of Ethics in the Development of Artificial Intelligence (AI) and Autonomous Systems (AS)

December 2016, IEEE http://standards.ieee.org/news/2016/ ethically_aligned_design.html Lethal Autonomous Systems and the Plight of the Non-combatant July 2103, Ronald Arkin, Georgia Institute of Technology http://www.cc.gatech.edu/ai/robot-lab/ online-publications/aisbq-137.pdf Campaign to Stop Killer Robots https://www.stopkillerrobots.org/

Sarah Underwood is a technology writer based in Teddington, U.K.

“Autonomous
systems could be
more unpredictable
than humans, which
in warfare could
lead to disastrous
consequences.”
ACM
Member
News

DETERMINING NORMS
FOR CYBER WARFARE
Patrick
McDaniel, a
Distinguished
Professor in the
School of
Electrical
Engineering
and Computer Science at
Pennsylvania State University
(Penn State), says that when he
was 11 years old, his father
brought home a TRS- 80 portable
computer from Radio Shack,
and handed him the manual to
BASIC. “Within 10 minutes I
was addicted, and I have never
looked back. I have bachelor’s,
master’s, and Ph.D. degrees in
computer science, and it has
never even been a thought to do
anything else.”
McDaniel obtained his
undergraduate degree at Ohio
University in 1989, and his
master’s degree at Ball State
University in 1991. He then
worked to develop some of the
first IP networking hardware as
a project manager at Primary
Access Corp. in San Diego,
which was acquired by 3Com
in 1995.

He later earned his Ph. D. in computer science and engineering at the University of Michigan, Ann Arbor. McDaniel spent several years as a senior research staff member at AT& T Labs in New Jersey, before joining the faculty at Penn State in 2004.

McDaniel is director of the Institute for Networking and Security Research at Penn State, and also university lead for the U.S. Army Cyber Security Research Alliance, a 10-year project to develop an understanding of how to make security-relevant decisions in cyberspace.

One area McDaniel is
focused on concerns the norms
for international cyber warfare.
He works to help define and
set standards for what is
allowable; in effect, a Geneva
Convention for cyber warfare.
“Right now, because nothing
is set up, it is really hard to go
to the U.N. Security Council for
sanctions when you haven’t set
up any norms.”
—John Delaney