Tuesday, February 10, 2004

Digital Confusion

Yeah, it's a Slashdot scrape, but it's important.

Digital forensics, especially image enhancement and incident tracing, are undergoing the same growing pains as did fingerprints and DNA. With digital forensics, it's that much more difficult as it's easier to fake ones and zeros than it is to fake molecular constructs. It's always an uphill climb for any technology to be used as scientific evidence in criminal cases.

Anyone see the problem in the following quote from the defense lawyer in the CNN story?

"Until there's a history of [what was done and when], not only will I attack it, it should be attacked. Otherwise, you are relying solely on the word of the person doing the work. That's not something I would like to do when someone's facing life in prison or death."

For those that don't see it, think about expert witnesses. WIth DNA or fingerprints, each side supports or attacks the evidence presented via an one or more expert witnesses. Often, jury decisions are based on which expert witness appeared to be more knowledgable, whether they actually were or not. (Hmm... It just occurred to me that this has a lot in common with those vendors that are able to convince management to buy a product even though you've been telling them for the last six months that the product is junk.) WIth digital evidence, until specific techniques become generally known and accepted as "common knowledge", we're going to see decisions like "a trojan did it!".

No comments:

Post a Comment