Following sharp increases in the first several years after 9/11, the total estimated costs of implementing the national security classification system seem to have leveled off at around $10 billion annually, according to a new report to the President (pdf) from the Information Security Oversight Office. The total cost of protecting classified information in government and industry last year was $9.85 billion, down slightly from $9.9 billion the year before, ISOO director William J. Bosanko reported.
The May 19 ISOO report also provided a striking reminder of the tide of secrecy that has silently and inexorably concealed previously public information in recent years.
When classification cost data were first reported in April 1994 (at the initiative of then-Rep. David Skaggs of Colorado), the CIA made the surprising claim that its classification costs were classified. This was understood by outside observers such as Sen. Daniel P. Moynihan as a familiar expression of CIA’s excess of zeal. No one seriously believed that it was a legitimate national security issue, particularly because the cost data involved were rough estimates, not actual expenditures. As if to confirm that assessment, the classification costs incurred by all other intelligence agencies were incorporated as a matter of course in the ISOO cost reports each year through 2005 (pdf).
But since 2006, the CIA’s silly secrecy has become prevalent in the intelligence community and other agencies have adopted the claim that their estimated classification costs are national security secrets too. In the latest ISOO report to the President, Mr. Bosanko noted that the cost estimates for CIA, DIA, ODNI, NGA, NRO, and NSA were all classified “in accordance with Intelligence Community classification guidance.”
“Understanding the financial costs associated with keeping information secret is essential to any effort to begin scaling back the scope of secrecy and making protection more efficient,” according to the 1997 Report of the Moynihan Commission (Chapter 1, page 9). Such an effort is naturally frustrated when classification itself is used to conceal those costs.
Americans are paying too much for almost everything, because the United States has long treated its trucking industry as an artifact to be preserved rather than as an opportunity for innovation.
These ideas aim to advance the detailed policy solutions needed to foster public trust and implement fairness in the adoption of AI across diverse domains, from healthcare and government benefits to rural access, education, and worker protections.
The evidence is clear: algorithmic pay-setting is established in app-based work, and payroll/timekeeping failures show how software can produce systemic wage harm at scale
While a few states have taken steps to implement decision-making mechanisms for certain AI systems, too many leaders are simply accepting narratives about AI’s purported public benefit at face value – jumping to the “how” of AI implementation before thoroughly vetting potential systems and deciding whether they are appropriate to use at all.