I have worked with software development and database programs, including data mining, enough to understand what data mining programs are capable of. I have also worked closely with systems administrators and architects for one of the US’ most secure networks. No doubt, the complexity of the NSA’s setup will be much more involved than those I have worked with in private industry and the non-secret information I had access to at the nuclear labs, yet the basic principles are the same—only more refined and greater in scale.
These programs mine for keywords and phrases, not for context or meaning. ...
They are quantitative versus a qualitative analysis of what a communication. I imagine there’s a review process which goes back to the original message and assesses whether the keywords actually are used in such a way as to mean something ominous and threatening.
Today I heard on messages he had sent overseas, is suing for any NSA intercepts of his client's messages overseas. During trial, al-Timini maintained that he had actually been preaching against terrorism, not for it. Since the revelation that the NSA was spying on US citizens, this man’s lawyer, Jonathan Turley, believes that these messages exist because in the government’s case against his client, there were numerous references to communications that could not be produced at trial because of national security.
Let’s assume for the moment that al-Timini is correct. Assume that his communications overseas contained such keywords as jihad, terrorism, etc. The NSA data mining programs picks up on these messages and forwards them to the appropriate next stage. Let’s assume, however, that the messages are not reviewed for content and meaning. As far as we know now, this might indeed be the case. In this scenario, someone is being arrested and tried because a computer program identified keywords—irregardless of meaning, context, nuance, etc. It is a purely quantitative analysis that has taken place.
I would hope, however, that the government’s case against this terrorist rested on more evidence than inaccessible communications overseas. I have to have faith that the jury decided the man’s guilt based on the evidence. Yet, I am not even sanguine about this possibility. As recent surveys of juries show, most jurors presume that because someone has been charged by the government they are therefore guilty—otherwise, the thinking goes, the government would not have charged the person in the first place.
What makes this case and cases like it more troubling, however, is the technical nature of the evidence. As numerous trials that include highly complex technology have shown, the average Joe in the street does not easily grasp the technicalities involved. Most people are not only intimidated by the technology, they are also simply incapable of grasping the concepts involved since they do not use the technology on a daily basis.
Of course, until the facts are known, all of this is purely speculative. Yet, as long as the NSA continues to withhold the nature of its spying efforts, such speculation will not only be warranted but also necessary—if only to force the government’s hand at clearing up the misconceptions and misunderstanding.
Unless, of course, they simply do not care—secure in their power, shrouded by obscurity and easily deflecting scrutiny or oversight by either the people’s representatives or the people themselves.
Updates/Related Articles[Prosecutor]Anthony Barkow acknowledged in his closing arguments to a jury ... "Mohammed Yousry is not someone who supports or believes in the use of violence."
Still, the prosecutor persuaded the jury to convict Yousry of supporting terrorism. Yousry now awaits sentencing in March, when he could face 20 years in prison for translating a letter from imprisoned Muslim cleric Omar Abdel Rahman to Rahman's lawyer in Egypt.
Tuesday, January 10, 2006
Computer Justice
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment