ISOO Asks Attorney General to Rule on Cheney’s Role
In an extraordinary internal challenge to the unruly Office of the Vice President (OVP), the Information Security Oversight Office (ISOO) has formally petitioned the Attorney General to direct the OVP to comply with a requirement that executive branch organizations disclose statistics on their classification and declassification activity to ISOO.
For the last three years, Vice President Cheney’s office has refused to divulge its classification statistics to ISOO, despite a seemingly explicit requirement that it do so. Prior to 2002, such information had routinely been transmitted and reported in ISOO’s annual reports to the President.
The disclosure requirement appears in ISOO Directive 1 (at section 2001.80): “Each agency that creates or handles classified information shall report annually to the Director of ISOO statistics related to its security classification program.”
Such ISOO directives “shall be binding upon the agencies,” President Bush wrote in Executive Order 13292 (section 5.1). Significantly, an “agency” here means not only a statutorily-defined executive branch agency (which would not include the OVP), but also refers to “any other entity within the executive branch that comes into the possession of classified information” (which would include the OVP).
Last May, the Federation of American Scientists urged ISOO to press for the Vice President’s compliance.
“Since the Office of the Vice President has publicly staked out a position that openly defies the plain language of the executive order, ISOO now has a responsibility to clarify the matter,” we wrote (pdf) at that time. “Otherwise, every agency will feel free to re-interpret the order in idiosyncratic and self-serving ways.”
This week ISOO indicated that it was actively pursuing the matter.
“With respect to the question you raised, I was unsuccessful in achieving a common understanding with OVP,” wrote ISOO director J. William Leonard in a February 5 email message.
“Accordingly, in early January, pursuant to section 6.2(b), Executive Order 12958, as amended, I wrote the Attorney General requesting that he render an interpretation on the issue,” he wrote.
(Section 6.2(b) of the executive order states that “The Attorney General, upon request by the head of an agency or the Director of the Information Security Oversight Office, shall render an interpretation of this order with respect to any question arising in the course of its administration.”)
“I have not received a reply to this request as of yet,” Mr. Leonard wrote.
He declined to provide a copy of his January letter to the Attorney General, explaining that it is pre-decisional.
The Justice Department has been asked at least once before to resolve a dispute over implementation of the executive order on classification.
In 1999, the Central Intelligence Agency refused to accept the jurisdiction of the Interagency Security Classification Appeals Panel over Agency classification activity. But the Justice Department Office of Legal Counsel issued a ruling in October 1999 that the CIA classification decisions were indeed subject to ISCAP review. That state of affairs was modified by President Bush in 2003, when he effectively gave the Director of Central Intelligence a veto over ISCAP decisions.
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.