Joe's Jottings

Jottings Number 69, by Joe Podolsky:

From: JOE_PODOLSKY@HP-PaloAlto-om4.om.hp.com

Date: Thu, 13 Feb 97 09:27:49 -0800

Subject: Good Business


One nice thing about my job is that I don't have to worry too
much about ethics.  Ethical behavior is pervasive at
Hewlett-Packard; we think about it only in the very rare
instances when something bad happens.  And, in those few cases,
the corporate reaction is generally more social than
bureaucratic; the error is quickly corrected, the offender is
treated appropriately, and the process usually is completed with
respect and privacy.  This is Washington, D.C. - NOT!  

Moreover, I'm a quality manager.  I've got a charter to yell
about any problems I see, ethical or otherwise.

But, thanks to the Internet and the explosion of electronic
commerce, all of us in information technology need to give more
thought to security, which leads us directly to ethical issues. 
W. Edwards Deming said, "In God we trust, but everyone else has
to bring data."  We can paraphrase that and say, "In God we
trust, but everyone else has to have a password."  

IT professional organizations have long been advocates of
ethics.   The International Federation of Information
Professionals (IFIP), a global association of organizations such
as the Association of Computing Machinery, recently published a
revised set of ethical guidelines.  These are discussed in the
January 1997 issue of the ACM SIGCHI Bulletin in an article by
John Karat and Clare-Marie Karat.   IFIP lists thirty codes,
grouped into these five categories: 

-  Respect (for individuals, for the public, for institutions,
for quality of life)

-  Personal (or institutional) qualities (e.g., honesty,
acceptance of responsibility, courage)

-  Information privacy and data integrity

-  Production and flow of information (e.g., information about
specifications and tests should be available to involved people,
even at the risk of violating the privacy guideline)

-  Attitude toward regulations (i.e., respect laws, regulations,
and professional standards)


The SIGCHI article points out there are broader (and grayer)
ethical issues not covered in the IFIP list, things like unequal
distribution of information and lack of respect for cultural
diversity.  My problem with the IFIP list, however, is that it
is like a set of laws in the Old West, with nary a sheriff for
miles around.  

Deborah G. Johnson is a professor of philosophy at Rennselaer
Polytechnic Institute.  She has written several books on
computer ethics, and she wrote an article for the January 1997
issue of _Communications of the ACM_ entitled, "Ethics Online." 
Johnson makes the connection between security and ethics.  She
says that neither ethical pronouncements nor technical fences
such as firewalls and encryption will be enough to ensure data
integrity and privacy.  She sees security as a moral issue. She
says that, "Our only hope is for individuals (online) to
internalize norms of behavior.  That is how most behavior is
controlled offline."

Johnson says that there are three special characteristics of
online communications that may affect moral behavior.  The first
of these is "scope."  "It seems, " Johnson writes, "to be the
combination of vastness of reach, immediacy, and availability to
individuals for interactivity that makes for something unusual
here... We might think of scope as power."

The second characteristic is "reproducibility."  The basic moral
problem here is that we can easily kid ourselves into thinking
that copying something is OK.  It's easy, cheap, and often
undetectable.  And the person we copied from still has the
original. 

The third, and perhaps most basic ethical characteristic is
"anonymity."  Johnson points out that, "... trust is difficult
to develop in an environment in which one cannot be sure of the
identities of the people with whom one is communicating."  She
observes that anonymity makes it harder to detect and catch
criminals, that it allows people to act without the normal
checks of social control, and that it creates doubts about the
source and, therefore, the integrity of shared information.

Johnson's proposed solution, is, to me, rather weak.  She
basically says that the online buyer must beware, that we must
try to apply the same standards of trust to cyberspace that we
would offline.  But, as she points out earlier, online anonymity
makes lying easy.  We have thousands of years of body language
that warns us about offline liars unless they are also good
actors.   But as the cartoon says, no one on the Internet knows
you're a dog.

In spite of Johnson's concerns about technological fences,
that's what we are basically using as the first lines of
defense.  In addition, systems for authentication,
authorization, and encryption are being developed to combat the
specter of anonymity.  We'll just use our skills with
information technology to protect our information technology
applications.  Hmmm.  That's like taking a shot of scotch as a
hangover cure.

My basic question is, "Are the ethical rules in cyberspace
different from those in the real world?"  In the real world,
people are expected generally to obey laws and follow social
norms.  Media and advertising have special "rules" to keep from
misleading us.  Some of those rules are fuzzy, as we learned in
the recent trial that found ABC liable when they had an
undercover reporter video tape bad food handling practices.

What are the appropriate ethical rules for IT professionals? 
The IFIP list shown above is pretty general.  Take, for example,
the issue of data integrity.  Should we have to analyze every
file for its statistical accuracy and place an appropriate
warning label on each report from that file?  

How concerned should we be about the use of information
technology?  For example, Cypress Semiconductor in San Jose uses
"killer systems" to automatically cut off various services from
people who don't meet certain task commitments.  How would you
feel if you were asked to build a system like this?

Where do we draw the line?   IT systems not only do what we want
them to, but they also invariably have unintended consequences. 
Vaughn Merlyn and Sheila Smith of the Omega Point Consulting
organization, wrote an article on this subject for the January
20, 1997 issue of _Computerworld's_ Leadership Series.  They
give two reasons for some of the unintended consequences.  They
say, first, that we assume "that information systems projects
are more logical, straightforward, and free of political
considerations than they really are... The second level of
failure is ... a lack of thought about the larger context." 
They say that we are often shortsighted in our ways of using
technology, not realizing or not communicating some of the
implications of the choices we make.

These "unintended consequences" are problems we see all the
time.  Merlyn and Smith talk mainly about business issues, such
as projects that are late and over budget and that cause various
types of lost opportunities.   When our customers ask for things
that may cause problems, how strongly should we, as experienced
IT people, voice our concerns.

And what about systems problems that are the types of "normal
accidents" that we discussed in the last jottings, the kind
caused by complexity that can disrupt vital services or cause
injury.

I don't know the answers to these questions, and I'd really like
your opinions on them.

I do believe, however, that security and ethics are two sides of
the same coin.  The privileges of trust and openness are
withdrawn when those precepts are violated, and we must then put
up electronic fences.  In application systems, the equivalent of
firewalls are strict standards and structured audits.  But
audits are generally most effective in business settings. 
Information technology is moving beyond the office into living
rooms.  If we have V-chips in our television sets, will we also
have to put their software equivalents in our Internet browsers?

Hewlett-Packard has continued to build its culture based on
moral beliefs of those of us following the examples set by Bill
and Dave.  But ethical strength is also good business. 
Information technology has no one set of role models on whose
beliefs we can build.  Instead, the on-going task is in our
hands.


Joe

Back to Joe's Jottings