Measure Twice, Shoot Once: Higher Care for CIA-Targeted Killing
Bibliographic record
Abstract
For almost a decade, the United States has deployed unmanned aerial vehicles, or "drones," to kill targeted members of Al Qaeda and the Taliban. Central Intelligence Agency (CIA) drone strikes in Pakistan have, in particular, stirred strong debates over the legality of such actions. Some commentators insist that these strikes are legal under international humanitarian law (IHL) or as a matter of self-defense. Others insist that the United States' targeted killing amounts to murder. It is critical for the law to determine how to control killer drones and the future of warfare. As technology evolves, drones will develop sharper senses and become more precise and lethal. The power to use drones to find and kill specific human targets-and states' temptation to use (and abuse) that power- will grow over time. On other fronts, drones may become fully automated, and their use in surveillance may spread along the borders with Canada and Mexico and into the U.S. heartland. To rein in the killer drones, this Article looks to foundational IHL principles to develop limits on the CIA's campaign in Pakistan and on the possible extension of that campaign to other countries outside the United States. In particular, this Article argues that IHL's requirements of distinction and military necessity generally require the CIA to achieve a very high level of certainty that a targeted person is a legitimate object of attack before carrying out a drone strike. To capture this level of certainty, one might borrow the "beyond reasonable doubt" standard from the criminal law, the "clear and convincing" standard from civil law, or create some new phrase. Also, to honor the principle of precaution, the CIA's Inspector General must review every CIA drone strike, including the agency's compliance with a checklist of standards and procedures for the drone program. The results of these reviews should be made as public as consonant with national security. These controls are, in the language of IHL, "feasible precautions" for the remote-control weapons of the new century. The Article closes by considering whether targeting of U.S. citizens by the U.S. government should be subject to stricter due process controls than targeting of non-Americans-a point that also has stirred controversy. The Article concludes that, if the controls on targeted killing are not good enough for U.S. targets, they are not good enough for Pakistanis, Yemenis, Somalis, and others. The law can develop a set of standards to ensure that states use targeted killing, whether as part of an armed conflict or in self-defense, only against legitimate targets- no matter their citizenship.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.013 | 0.037 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.020 | 0.011 |
| Scholarly communication | 0.012 | 0.010 |
| Open science | 0.004 | 0.021 |
| Research integrity | 0.011 | 0.020 |
| Insufficient payload (model declined to judge) | 0.035 | 0.008 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".