Algorithmic accountability is usually framed as a transparency problem. If affected people could understand how the algorithm reached its decision — the features it weighted, the thresholds it applied — they could contest unfair outcomes. This framing has produced a decade of research on explanations: feature importances, counterfactual examples, attention visualizations.
Note (2026-11-13; read against the paper's full text): the essay presents the 9 percent and 97 percent figures as a stark effect of evidentiary access. In the paper the 9 percent is 1 success in 11 resolved cases with no access, and the 97 percent is 56 of 58 resolved partial-access cases in domains without immunity shields; full-access cases succeeded 88 percent of the time, because all six platform cases had full scrutiny and were denied under Section 230 immunity, and 48 of the 168 cases were still ongoing and excluded from the rates. The paper states that the association almost certainly reflects selection effects, that its empirical contribution is diagnostic rather than causal, and that the data do not show whether providing access would change outcomes for excluded cases; its normative argument rests on procedural fairness. The essay's claim that the low success rate is a 'gap in power' is the essay's reading. The paper also introduces a taxonomy of contestation failures in which explanation addresses one (opacity) and four others are left unaddressed, which the essay omits.
Stewart (arXiv:2603.22716) analyzed 168 legal cases challenging algorithmic decisions and found that transparency is the wrong axis. The cases revealed a two-gate structure: first, procedural access to evidence about how the system actually works; second, substantive liability rules that determine whether the evidence matters.
The numbers are stark. Cases that lacked evidentiary access succeeded 9% of the time. Cases with evidentiary access succeeded 97% of the time in domains without liability protections. The bottleneck isn't understanding — it's access. You can explain a decision perfectly, but without the ability to test the system with modified inputs, to verify that the explanation matches the actual decision process, the explanation is unfalsifiable.
The proposed solution is counterfactual interrogation rights: the ability to submit modified inputs and observe changed outputs, without requiring disclosure of the model itself. This is a black-box audit right — testing the system's behavior without opening it. The accountability shifts from “explain how you decided” to “prove your decision would survive if the protected characteristic were different.”
The through-claim: accountability that offers explanation without evidence is procedurally decorative. The 9% success rate without access isn't a gap in understanding — it's a gap in power. Transparency without testability is a wall painted to look like a window.