By
Gary Milhollin
New York Times
June 16, 2000, p. A 33
The Los Alamos National
Laboratory is facing another nuclear mystery. Two computer hard drives
-- full of weapons secrets -- are missing. There is no proof that they
have fallen into the wrong hands, but even if they have simply been misplaced,
there is cause for worry. The missing drives are but a symptom of a widespread
problem: sensitive information about nuclear weapons is not protected
as carefully as nuclear hardware, even though it is just as dangerous
if left unsecure.
The Energy Department,
which oversees the national laboratories, isn't revealing what the drives
contain. But the data is used by the department's Nuclear Emergency Search
Team, the experts who must be ready, at a moment's notice, to jump on
a plane and find any nuclear bomb that might be planted on American soil
-- and then disable it.
The data on the
drives, each about the size of a deck of cards, tells the team how to
detect nuclear material in a multitude of situations -- when it is underground,
or encased in concrete or masked by the presence of natural or manmade
background radiation.
"What you have is
an entire library of nuclear signatures," a Pentagon official familiar
with the team's operations told me. These are the guides that help the
team uncover various kinds of nuclear material.
What worries this official is that the data could teach a foreign bomb designer ways to evade American detection efforts.
"If you know what
your enemy is looking for, it is much easier to hide it," he says. "If
a mountain lion could morph his footprints into a turkey's, think how
difficult lion hunting would be. You would never know where the lion might
be hiding."
The data could be
used to foil our efforts to disable a bomb. Some terrorist could "put
an extra wire, or an extra bump, or an extra piece of metal where there
isn't supposed to be one," he warns. This could prevent the search team
from figuring out what kind of bomb they were dealing with.
The missing data
also reveal how a stolen bomb might be set off. Most nuclear weapons are
prevented from exploding by an internal safety system, but the data would
help bypass it. To make matters worse, the drives include information
about Russian nuclear designs. A terrorist group or rogue nation with
the drives is a chilling thought.
The gaps in security
that produced this mess are dismaying. They reveal the government's failure
to grasp an essential point: Information can be more important than hardware.
As terrifying as the theft of one or two bombs would be, it is not nearly
as frightening as the loss of the formula for producing an unlimited number.
Yet the protection of information does not get the same respect, and that extends beyond weapons labs.
The State Department,
for example, has had its own security problems, including the disappearance
of a laptop that contained highly sensitive files, including intelligence
sources and methods related to weapons proliferation. And John Deutch,
the former director of Central Intelligence, loaded classified information
into unsecure computers in his home.
In the case of the
missing hard drives at Los Alamos, a total of 86 members of the Nuclear
Emergency Search Team had access to the vault in the X division, where
the hard drives were stored. Of them, 26 could go in without escorts,
and remove the hard drives without signing out or leaving any record that
they had taken the material.
Thus, if two of
the team members had not discovered the hard drives missing on May 7,
when they searched the vault as a forest fire raged near the laboratory,
the missing drives might have gone undetected even longer.
These casual security measures are a far cry from the stringent procedures used to monitor nuclear material. The United States protects its nuclear weapons zealously and tracks the plutonium and uranium used to make them as best it can. We try to follow the movement of every gram of nuclear material and to verify its whereabouts continuously. Shouldn't we start protecting our nuclear information with the same resolve?