
What's behind the rash of university data
breaches?
|
By Jay Cline |
|
Purdue University last month reported its seventh data
breach in the past four years. But Purdue is hardly alone.
According to my records, over 300 publicized privacy incidents
have occurred at U.S. institutions of higher learning since
2001, with at least 53 colleges and universities experiencing
multiple breaches (see table at end of article).
The regular stream of university data-breach reports has
prompted Adam Dodge, assistant director for information security
at
Eastern Illinois University, to devote a blog —
Educational Security Incidents — to the topic.
When I last covered the issue four years ago (see
"Security breaches challenge academia's 'open society' "),
universities were the leading sector for publicized breaches.
The same is true today.
What's going on? Why haven't things changed?
John Correlli of Los Angeles-based
JMC Privacy Consulting Group has some answers. Correlli
recently published a detailed analysis of the topic,
"Breaches in the Academia Sector." Correlli identifies the
top three root causes of university breaches: unauthorized
access, usually inside jobs; accidental online exposures; and
stolen laptops.
"Privacy governance in academia is far too frequently thrown
into the laps of the IT folks, who are then told, implicitly or
explicitly, that privacy isn't a priority until it's a problem,"
Correlli told me.
Correlli also points to unique threats and vulnerabilities in
academia:
-
The open nature of the university physical and technical
environment.
-
Department fiefdoms inhibiting central policy
enforcement.
-
A customer user population that is relatively low paid,
lives "on site" and experiences high turnover.
There is some debate over whether students perpetrating
intentional breaches or staff making unintentional data
disclosures are the principal source of data risk within
universities. I think both are worth monitoring, but would pay
special attention to students. Why? Twice a year, college
students are under extreme duress to produce results that their
futures depend on. The statistics appear to bear this out.
Looking at the months of the reported breaches, peak activity
occurs during the traditional finals weeks of fall and spring
semesters. In contrast, the fewest breaches are reported during
months when students aren't around (see graph).
Elevated data risk during finals week?
A monthly breakdown of university data breaches reported
since 2001 shows January and May as the peak months.
Allowing for a few weeks to detect and report these
incidents, the actual peak in incident activity may be
occurring during the final weeks of the fall and spring
semesters. Number of reported breaches at universities,
by month:

Susan Blair, chief privacy officer at the
University of Florida, generally agrees with Correlli. In a
presentation she shared with me, Blair lists these as the top
reasons for university breaches:
-
Data-rich information systems creating a natural target.
-
Outdated and nonenforced data-security safeguards.
-
Sophisticated intruders, with potential criminal intent.
-
Careless or inattentive data systems management.
-
Negligent hiring practices or employee misuse of data.
-
Demonstrated opportunities for repeat access.
-
Business partners or research sponsors who fail to
protect information.
"The typical academic network is a maelstrom of collaborative
activities that generally precludes the kind of restrictions
that a corporate network would impose," said Michael Corn, chief
privacy and security officer at the
University of Illinois at Urbana-Champaign. "We accept this
risk as a precondition for academic endeavors.
"Universities are uniformly more forthcoming when data
breaches occur due to a culture of transparency in these
matters," Corn added.
Rodney Petersen, government relations officer and security
task force coordinator at Washington-based
EduCause, also believes there is a reporting bias that
overestimates the data risk in academia. "It is not fair to
conclude that higher-education environments are any less secure
than their government or corporate counterparts," he told me.
"Institutions of higher education have been disclosing security
breaches long before they were required to do so under
individual state laws because institutional officials err on the
side of protecting their students, faculty and alumni.
"Corporations may be far more circumspect before deciding to
report incidents because of concerns about consumer confidence
or impact on shareholder value," he added.
Rachel Krinsky, assistant director of compliance and privacy
at the
University of Connecticut, agreed with Peterson. "Many
universities are large and made up of multiple colleges,
campuses and divisions. As a result, some universities have
decentralized networks and systems without a centralized
oversight function to monitor them in the same way as may be
done in other sectors," she added.
"This means that a university may have multiple networks and
systems to contend with," Krinsky continued, "and each one is
managed differently and separately."
What's the outlook for data privacy in academia?
Several university privacy and security leaders told me off
the record that the role of the chief privacy officer needs to
be elevated in academia before major progress can be made.
Indeed, in a sector regulated by the
Health Information Portability and Accountability Act (HIPAA),
Family Educational Rights and Privacy Act (FERPA),
Gramm-Leach-Bliley Act (GBLA),
Fair and Accurate Credit Transactions Act (FACTA) ID Theft Red
Flags Rules,
Payment Card Industry Data Security Standard (PCI DSS, and
state-level laws on Social Security numbers and breach
notification, it's surprising how few CPOs there are in
academia. I was able to find just 20 to contact for this
article.
More will certainly be found attending the
Academic Medical Centers Privacy and Security Conference,
International Association of Privacy Professionals Privacy
Summit, and EduCause/Internet2 Security Professionals
Conference over the next two months.
But until university trustees and boards of regents fund more
robust privacy programs and hold university presidents more
accountable for their privacy status, don't expect another
sector to overtake the lead in the reported-breach column.
Double trouble
Over 50 colleges and universities have experienced
multiple reported privacy incidents since 2001. At a
state level, California is home to seven doubly breached
universities, while Ohio follows at four schools. At
least eight universities have experienced four or more
publicized privacy incidents.
| University |
Dates of reported incidents |
| Austin Peay State (Tennessee) |
December 2008, July 2005 |
| Cal Poly (California) |
December 2008, August 2006, July 2005 |
| California State – Dominguez Hills (California) |
March 2006, July 2005 |
| California State – Stanislaus |
January 2008, May 2006, August 2005 |
| Carnegie Mellon (Pennsylvania) |
October 2007, April 2005 |
| City University of New York |
November 2007, September 2005 |
| Duke University (North Carolina) |
December 2007, May 2005 |
| East Carolina University (North Carolina) |
February 2007, June 2005 |
| Florida International University |
May 2006, April 2005 |
| Georgetown University (District of Columbia) |
January 2008, March 2006 |
| Georgia Tech University |
June 2007, February 2007, November 2005, March
2003 |
| Harvard University (Massachusetts) |
November 2008, March 2008 |
| Indiana University |
November 2005, February 2001 |
| Iowa State University |
December 2005, July 2005 |
| Kansas State University |
January 2009, November 2007 |
| Kent State University (Ohio) |
September 2005, June 2005 |
| Michigan State University |
July 2005, April 2005 |
| Middle Tennessee State University |
February 2008, May 2005 |
| Montana State University |
November 2007, October 2007, December 2006 |
| New Mexico State University |
January 2008, April 2007 |
| Northwestern University (Illinois) |
June 2007, May 2007, July 2006 |
| Ohio University |
December 2008, June 2006, May 2006, May 2006,
April 2006 |
| Ohio State University |
December 2008, May 2008, April 2007 |
| Oklahoma State University |
April 2005, February 2001 |
| Purdue University (Indiana) |
February 2009, September 2007, July 2007, April
2007, September 2006, April 2006, May 2005 |
| Stanford University (California) |
June 2008, May 2005 |
| Tennessee Tech University |
January 2008, September 2007 |
| Texas A&M University |
November 2008, February 2008, June 2007 |
| University of Akron (Ohio) |
January 2008, October 2007 |
| University of Alabama |
February 2009, June 2006 |
| University of California, Los Angeles |
December 2006, April 2004 |
| University of California, Berkeley |
May 2006, March 2005 |
| University of California, Davis |
June 2007, March 2005 |
| University of California, San Francisco |
May 2008, April 2007, March 2005 |
| University of Colorado |
April 2008, May 2007, December 2006, August 2005 |
| University of Delaware |
May 2006, January 2006, November 2005 |
| University of Florida |
February 2009, January 2009, November 2008, June
2008, May 2008 |
| University of Georgia |
January 2008, September 2005, January 2004 |
| University of Idaho |
March 2007, January 2007 |
| University of Iowa |
October 2007, June 2007, September 2006, July
2006, May 2005 |
| University of Kansas |
September 2007, January 2006, April 2004 |
| University of Kentucky |
August 2006, August 2006, June 2006, June 2006 |
| University of Michigan |
September 2007, July 2007 |
| University of Nebraska |
July 2008, February 2007, March 2006 |
| University of New Mexico |
April 2007, January 2007 |
| University of San Diego (California) |
December 2005, July 2005, January 2005 |
| University of South Carolina |
June 2008, September 2007, August 2006 |
| University of Tennessee |
July 2006, October 2005 |
| University of Texas at Austin |
November 2004, March 2004, March 2003 |
| University of Toledo (Ohio) |
April 2008, August 2007 |
| University of Utah |
June 2008, August 2005 |
| University of Virginia |
April 2008, June 2007, November 2006, April 2006 |
| Virginia Commonwealth University |
December 2006, September 2006 |
Source: Minnesota Privacy Consultants
This article originally appeared in
COMPUTERWORLD on
9 March, 2009.
Other articles by Jay Cline |
|
|