Databases Reference
In-Depth Information
It would seem it enables one to prove later in court that (at least a copy of)
the Work was in one's possession at the time of release. In the following we
address these and other related issues.
2.1 Rights Protection through Assessment
The ability to prove/assess rights convincingly in court constitutes a deterrent
to malicious Mallory. It thus becomes a tool for rights protection if counter-
incentives and legal consequences are set high enough. But because Informa-
tion Hiding does not provide a means of actual access control, the question of
rights protection still remains.
How
are rights protected here?
It is intuitive that such a method works only if the rightful rights-holder
(Alice) actually knows about Mallory's misbehavior
and
is able to prove to
the court that: (i) Mallory possesses a certain Work
X
and (ii)
X
contains a
“convincing” (e.g., very rare with respect to the space of all considered similar
Works) and “relevant” watermark (e.g., the string “(c) by Alice”).
What watermarking itself does not offer is a direct deterrent. If Alice does
not have knowledge of Mallory's illicit possession of the Work and/or if it
is impossible to actually prove this possession in court beyond reasonable
doubt, then watermarking cannot be deployed directly to prevent Mallory. If,
however, Information Hiding is aided by additional access control levels, it
can become very effective.
For example, if in order to derive value from the given Work (e.g., watch a
video tape), Mallory has to deploy a known mechanism (e.g., use video player),
Information Hiding could be deployed to enable such a proof of possession, as
follows: modify the video player so as to detect the existence of a watermark
and match it with a set of purchased credentials and/or “viewing tickets”
associated with the player's owner. If no match is found, the tape is simply
not played back.
This scenario shows how watermarking can be deployed in conjunction
with other technologies to aid in managing and protecting digital rights. In-
tuitively, a certain cost model is assumed here: the cost of reverse engineering
this process is far higher than the potential derived illicit gain.
This illustrates the game theoretic nature at the heart of the watermarking
proposition and of information security in general. Watermarking is a game
with two adversaries, Mallory and Alice. At stake lies the value inherent in
a certain Work
X
, over which Alice owns certain rights. When Alice releases
X
, to the public or to a licensed but potentially un-trusted party, she deploys
watermarking for the purpose of ensuring that one of the following holds:
•
she can always prove rights in court over any copy or valuable derivate of
X
(e.g., segment thereof)
•
any existing derivate
Y
of
X
, for which she cannot prove rights, does not
preserve any significant value (derived from the value in
X
)
