DIGITAL SIGNATURE LEGISLATION: SOME REASONS FOR CONCERN
by Brad Biddle <biddle@acusd.edu>
February 20, 1996
[Copyright 1996 by Brad Biddle; permission granted for non-commercial
electronic redistribution]
A recent flurry of state digital signature legislation should provoke some
concern among consumer activists, privacy advocates, and others interested
in the evolving legal landscape of cyberspace.
At least ten states are developing or have already implemented digital
signature legislation. Much of this legislation is based on the pioneering
Utah Digital Signature Act, passed in 1995 (and currently in the process
of being amended). States which have introduced legislation based on the
Utah Act include Arizona, Georgia, Hawaii, Oregon, and Washington.
California has passed a different form of digital signature legislation,
and a bill is pending in Illinois which is similar to the California
approach. Florida and Virginia have formal resolutions pending which call
for legislative investigation into digital signature laws.
The American Bar Association's Information Security Committee (a
sub-committee of the Section of Science and Technology) released its
Digital Signature Guidelines in October of 1995. These Guidelines are
"general, abstract statements of principle" and are not intended as model
legislation. The Information Security Committee intended to release model
digital signature legislation in June of 1995, but this effort has been,
as one report describes it, "stymied by bureaucratic maneuvering."
(Information Law Alert, 10/13/95). In the absence of this model
legislation, the Utah Act has become a de facto model act.
The intent of this message (distributed to the "ca-digsig" mailing list
and to some other folks via direct e-mail) is to raise some concerns about
the Utah Digital Signature Act and its progeny. The author of this message
is a second-year law student at the University of San Diego School of Law,
and is writing an article on this topic for the San Diego Law Review.
Feedback and criticism are very welcome, and will likely be incorporated
into the developing article.
THE UTAH DIGITAL SIGNATURE ACT
[Sec. 46-3-101 et seq., Utah Code Annotated 1953]
No attempt will be made here to explain or summarize the complex and
detailed provisions of the Act. Generally, the Act envisions an
infrastructure in which computer users utilize state-licensed
certification authorities, online databases called repositories, and
public-key encryption technology in order to "sign" electronic documents
in a legally binding fashion. The Utah Act sets out an ambitious legal and
regulatory framework intended to implement a public key infrastructure.
It also carves out a place for digital signatures in the broader legal
landscape. That is, it provides digital signatures with legal status as
valid signatures and addresses a variety of issues relating to the place
of electronic documents in contract and evidence law.
Much of what the Utah Act accomplishes is laudable, and demonstrates how
legislation can effectively solve unsettled issues in the novel arena of
cyberspace. However, several aspects of the Act are troubling. The
potential problem areas can be categorized generally as liability,
privacy, and costs. A very brief discussion of each of these problem areas
follows.
LIABILITY
The Utah Act makes two policy choices concerning liability allocation
which are potentially troubling. First, consumers who participate in the
infrastructure developed under the Utah Act subject themselves to a far
greater risk of extensive liability than they face in other electronic
transactions, such as credit card or debit card transactions. Most
electronic transactions made by consumers are subject the Electronic Funds
Transfer Act (EFTA) which limits consumer liability in the event of fraud
to (in most cases) $50. Even if a consumer is negligent, liability is
still capped at a rather low fixed amount. Critics of this scheme argue
that it is paternalistic and ultimately drives up costs for other
consumers who are careful to avoid exposing themselves to fraud.
Supporters argue that if consumers were exposed to potential unlimited
liability when engaging in electronic transactions they would not
participate in these transactions at all, and the potential benefits of
electronic transactions would not be achieved. Also, supporters say,
consumers are often unable to prevent fraud, and forcing consumers to
prove they were not negligent anytime fraud occurs would be an
unreasonable burden.
Under the Utah Act, consumers are held to a negligence standard in
guarding their private encryption key. Thus, if a criminal obtains a
consumer's private key and commits fraud, the consumer is financially
responsible for that fraud unless the consumer can prove that the consumer
used reasonable care in guarding the private key. If the consumer cannot
prove this in court, or if the consumer was in fact negligent, then the
consumer will bear all losses resulting from the fraud. The arguments in
support of the EFTA may be applicable here. Will consumers participate in
a system which subjects them to unlimited liability? Is it sensible to
make consumers prove the absence of negligence?
(Two related points are worth noting. First, drafters of the Utah Act
initially advocated a strict liability standard, rather than negligence,
for the security of private keys -- even "worse" for consumers -- and the
drafters continue to advocate strict liability as an alternative for other
state legislators considering digital signature laws. Second, a plausible
argument can be made that the federal EFTA should preempt the state
digital signature legislation on this issue -- this question is
unsettled.)
There is a second troubling policy choice relating to liability. The Utah
Act limits the potential liability of one actor in the infrastructure --
the certification authority -- to a fixed amount (termed a "suitable
guarantee" and determined by a complex formula or by administrative rule).
This amount may be less than the actual damages a certification authority
can cause. This policy decision, designed to create certainty for an
entrepreneur contemplating a certification authority business and foster
development of a certification authority industry, may have unintended
consequences. It is easy to envision a scenario in which a certification
authority's private key is compromised -- by brute force cryptanalysis,
bribery, or incompetence, for example. A criminal with a certification
authority's private key could cause an immense amount of financial damage,
imposing huge losses on a number of innocent parties. These innocent
parties would be unable to recover their full losses from the
certification authority if the total of these losses was greater than the
amount of the "suitable guarantee" -- even if the certification authority
was totally at fault in creating the circumstances that led to the losses.
Because the certification authority would not have to bear the full costs
of any losses resulting from a compromised private key, they may not have
the incentive to take expensive precautions to protect against that
occurrence.
PRIVACY
The system contemplated by the Utah Act also raises several different
types of privacy-related concerns. At a broad level, one commentator has
pronounced the general type of system embodied in the Utah Act a "cultural
misfit" because every merchant and consumer potentially must register
with an outside authority in order to acquire the basic capacity to
transact commerce. In light of the more limited scope of the Utah Act and
the current state of electronic commerce, however, this argument is not
particularly persuasive.
More significantly, under the Utah Act's approach certain entities -- the
online databases of public encryption keys termed "recognized
repositories" -- will have unrestricted access to valuable
transaction-generated information that could expose sensitive
relationships among individuals or businesses. If Company A sends a
digitally signed message to Company B, Company B must verify the digital
signature by connecting to a state-recognized privately-managed database,
verifying the digital signature and making sure that Company A's
certificate is not on a certificate revocation list. This process, of
course, will leave electronic footprints. Could the owner of the
recognized repository disclose the fact that A and B were corresponding?
What if A and B were discussing a possible merger, or other transaction
with significant consequences in the securities markets? Similarly, could
the owner of the repository disclose to Joe Whistleblower's
defense-contractor employer that Whistleblower was verifying digital
signatures of a reporter from the New York Times? The Utah Act is totally
silent on this issue.
Additionally, the public databases contemplated by the Act could expose
financial data, information about affiliations, and other private
information to public scrutiny -- and put this information into the direct
marketing universe. Publicly-accessible certificates will contain the name
of subscribers and a "recommended reliance limit," a dollar figure that
may be a good indication of general financial standing. Certificates may
also indicate an individual's affiliation with a company or other
organization. There is no provision for anonymous or pseudonymous
certificates. Proponents of the Utah Act point out that participation in
the system established by the Act is voluntary, and that non-licensed
certification authorities will be available for individuals who object to
the requirements of the Act. However, in light of the advantages the Act
gives to licensed certification authorities (liability limitations,
presumptions concerning the legality of digital signatures, and the like)
this may not in fact turn out to be true. Additionally, some individuals
may be forced to use certificates in the course of their employment. Would
an employee who did not want to be listed in an easily searchable database
(perhaps because they were being harassed) be forced to quit his or her
job?
Finally, a very important privacy-related issue that is purposefully not
addressed in the Utah Act concerns whether the infrastructure contemplated
by the Act will support confidentiality of messages as well as legally
binding digital signatures (a technically feasible proposition, but a
politically sensitive one). The Utah Act empowers an administrative agency
to determine which public key encryption algorithms are appropriate. A
public key algorithm like RSA can be used both for encryption and digital
signatures. A public key algorithm like DSA (implemented in DSS) can only
be used for digital signatures -- it cannot be used to encrypt messages.
Should such a fundamental policy decision be made in the obscurity of an
administrative agency's rulemaking process?
COSTS
The Utah Digital Signature Act also raises several issues relating to
costs. The institutional overhead associated with creating and maintaining
the Act's infrastructure will be passed along to participants, and
participants must have access to expensive computer hardware and software
in order to participate in the system. One issue not addressed by the Utah
Act is whether citizens who are unable to afford these costs should be
provided with subsidized or reduced-cost access to th infrastructure. A
prominent commentator has noted that, in the long term, the type of system
embodied in the Utah Act is "anticipated to become indispensable for
conducting government, business, and even private affairs."
Another cost-related issue concerns the costs associated with legislative
endorsement of one particular technology (public-key encryption
technology, or more narrowly, specific implementations of this technology)
and whether this endorsement will affect the development of alternative
solutions to the problems posed by communications over open computer
networks. An advocate of a particular biometric technology has argued that
the type of infrastructure contemplated by the Utah Act is costly
overkill, and is far more complex and expensive than is necessary. Even if
one accepts the appropriateness of a public-key approach, note that costs
could vary widely depending upon which particular proprietary encryption
algorithms are licensed.
As originally passed, the Utah Act limited the role of certification
authority to Utah-licensed attorneys, financial institutions, title
companies, and government agencies. This sort of oligopolistic arrangement
is, of course, anathema to a vibrant, competitive market which would drive
down costs for consumers. The pending amendments to the Utah Act eliminate
this requirement. Some of the states which are following the Utah Act as a
model have retained this limitation, however.
CONCLUSION
Legislative activity concerning digital signatures is generally
appropriate and potentially helpful. The Utah Digital Signature Act,
particularly its provisions establishing the legal status of digital
signatures, is a step in the right direction. However, lawmakers
contemplating digital signature legislation should reconsider some of the
policy choices made by the Utah Act.
-------------------------------------------------
Brad Biddle, Legal Intern <biddle@acusd.edu>
Privacy Rights Clearinghouse, Ctr for Public Interest Law
http://pwa.acusd.edu/~prc
[The views expressed in this article are not necessarily those of the
Privacy Rights Clearinghouse or the Center for Public Interest Law.]
--- end forwarded text
--------------------------------------------------
The e$ lists are brought to you by:
Making Commerce Convenient (tm) - Oki Advanced Products - Marlboro, MA
Value-Checker(tm) smart card reader= http://www.oki.com/products/vc.html
Where people, networks and money come together: Consult Hyperion
http://www.hyperion.co.uk info@hyperion.co.uk
See your name here. Be a charter sponsor for e$pam, e$, and Ne$ws!
See http://thumper.vmeng.com/pub/rah/ or e-mail rah@shipwright.com
for details...
-------------------------------------------------