In Freedom of Information Act litigation, courts will almost always defer to a government agency when it asserts that national security requires that certain information remain classified. Judges say they are reluctant to “second guess” agency national security experts, and there is a substantial body of case law that discourages them from doing so.
But earlier this month, Judge Richard W. Roberts of the DC District Court considered an agency’s national security claim, found it unpersuasive, and rejected it (pdf).
In that FOIA case, Center for International Environmental Law vs. the Office of the United States Trade Representative, the plaintiff sought a USTR document concerning the U.S. negotiating position on the Free Trade Agreement of the Americas. USTR said the document was classified because the international negotiations were confidential and the document’s disclosure would result in damage to U.S. foreign relations.
“USTR argues that release of document 1 would constitute a breach of its agreement with the other nations participating in the FTAA negotiations. [USTR] states that [t]here is an understanding among the 34 participating governments, consistent with longstanding practice in multiparty trade negotiations, that they will not release to the public any negotiating documents they produce or receive in confidence in the course of the negotiations unless there is a consensus among the 34 governments to do so.”
But remarkably, the judge didn’t buy it, particularly since it was a question of releasing a U.S. document, not a foreign document.
USTR “has not shown it likely that disclosing document 1 would discourage foreign officials from providing information to the United States in the future because those officials would have no basis for concluding that the United States would dishonor its commitments to keep foreign information confidential,” he concluded.
“Although a court must defer to agency affidavits predicting harm to the national security, ‘[d]eference… does not mean acquiescence’,” Judge Roberts wrote. See the April 12, 2011 Memorandum Opinion here.
The ruling that international negotiations cannot necessarily be used as a pretext for classifying U.S. government information may have important ramifications in other policy areas.
So, for example, the U.S. government currently makes less information about the makeup of the U.S. nuclear arsenal under the New START Treaty than it previously did under the START regime, observed Hans Kristensen of FAS last month.
Although such stockpile information is generated and is regularly exchanged with the government of Russia under the provisions of New START, it is currently classified and has still not been made publicly available. If it became necessary to challenge the classification of this information in court, then Judge Robert’s new ruling might offer an apt precedent.
“Although the Constitution permits the judiciary to play a role in judging government secrecy claims and Congress has repeatedly endorsed that role, most prominently in the Freedom of Information Act, judges have been reluctant to question Executive Branch secrecy,” observed Meredith Fuchs in a 2006 law review article that argued for a more active judicial role in reviewing classification decisions. “Without judicial intervention…, the incentives on the Executive Branch to overreach far outweigh any checks on excessive secrecy.”
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.