The U.S. government’s approach to classifying and declassifying information needs a complete overhaul involving major technological upgrades according to a report by the Public Interest Declassification Board.
The current system was established in 1953 and has agencies classify information as top secret, secret or confidential. The board suggests creating two categories: top secret and “lower level.” Available technologies such as context accumulation, predictive analytics and artificial intelligence, should be piloted.
“Current page-by-page review processes are unsustainable in an era of gigabytes and yottabytes,” the report states. “New and existing technologies must be integrated into new processes that allow greater information storage, retrieval, and sharing. We must incorporate technology into an automated declassification process.”
In the current system, classifiers often struggle to distinguish what is confidential or higher secret classification and often are “erring on the side of protection.”
“There are many explanations for over-classification: most classification occurs by rote; criteria and agency guidance have not kept pace with the information explosion; and despite the Presidential order to refrain from unwarranted classification, a culture persists that defaults to the avoidance of risk rather than its proper management,” the report states.
If there isn’t a dramatic improvement in the declassification process, there will an exponential growth in the archival backlog of classified records awaiting declassification, and public access to the nation’s history will deteriorate further, the report stated.
New technologies should be tested through a series of pilot projects, which will help to discover, develop and deploy technology that will:
- Automate and streamline classification and declassification processes, and ensure integration with electronic records management systems.
• Provide tools for preservation, search, storage, scalability, review for access, and security application.
• Address cyber security concerns, especially when integrating open source information into classified systems.
• Standardize metadata generation and tagging, creating a government-wide metadata registry. Lessons learned from the intelligence community will be helpful here.
• Accommodate complex volumes of data (such as email, non-structured data, and video teleconferencing information).
Federal agencies, excluding the CIA and other intelligence agencies — spent $11.4 billion to protect secrets, more than double the amount a decade ago, according to a report by the Information Security Oversight Office. The total number of classification decisions quadrupled to more than 96 million from 2008 to 2011.
“We want to push the government out of its comfort zone to try to think through a new way of looking at this,” said Nancy Soderberg, the chairman of the Public Interest Declassification Board.