Archive Audit Suggests Overclassification is Rampant
A large fraction of the documents that were withdrawn from public access at the National Archives on purported national security grounds over the past several years did not meet the standard for classification and should not have been removed, according to an official audit of the activity released yesterday.
“This audit identified a significant number of withdrawal actions for classification purposes as inappropriate. Of the records sampled to date, 24 percent were clearly inappropriate and 12 percent were questionable.”
While focused on historical documents at the National Archives, the audit serves in effect as a snapshot of classification activity throughout the government, and it implies that a sizeable fraction of agency classification actions have no legitimate national security basis.
“To be effective, the classification process is a tool that must be wielded with precision,” said William Leonard (pdf), director of the Information Security Oversight Office, which performed the audit at the direction of Archivist Allen Weinstein.
“It is disappointing to note, as indicated by the sample contained in this audit, that even trained classifiers, with ready access to the latest classification and declassification guides, and trained in their use, got it clearly right only 64% of the time in making determinations as to the appropriateness of continued classification,” Mr. Leonard said.
“The damage such practices can inflict on the integrity of the classification system cannot be denied,” he said.
At a time when the Bush Administration is prosecuting even the receipt of classified information, and Members of Congress are seeking new measures to penalize leaks, the new data on overclassification tend to undermine the very premise of such actions.
Archivist Weinstein and Mr. Leonard of ISOO announced a series of steps to address the immediate issue of document withdrawal at the Archives as well as the larger issue of overclassification and misclassification.
“I am writing to all agency heads asking for their personal attention in ensuring that all of us engaged in advancing our country’s security perform our duty to ensure the highest effectiveness of this critical national security tool (i.e. classification),” Mr. Leonard said.
He said that several of the existing provisions in the executive order and implementing directive on classification could help to mitigate classification errors, including: challenges to classification, sanctions for unwarranted classification, and audits of classified collections.
“They just haven’t been used,” he said.
When properly structured — with specific numeric targets, secured financial obligations, independent monitoring, and meaningful enforcement — CBAs transform data center deals into durable community partnerships.
Protecting the public from the tech industry’s predatory business models and the next wave of AI harms is an enormous challenge, but we have the evidence that trying to build a healthier digital culture is absolutely worth the effort.
Opaque and insufficiently tested tools are increasingly shaping student outcomes without consistent transparency, civil rights review, or technical safeguards. States and the U.S. Department of Education can address these risks using procurement and oversight tools already within their authority.
Commercial artificial intelligence tools have recently emerged that are able to produce police reports. If the resulting reports are inaccurate, incomplete or biased, or if the process leaks confidential information, this could undermine the criminal justice system and harm citizens.