Contribution to the handbook
Priority · Assessed 2026-10-27
It surfaces consequential implementation limits for a possible safeguard: understandability requirements may trade off against sought-after benefits, encounter resistance, or be constrained by secrecy claims. These are qualified arguments rather than measured findings, but they matter to practical design.
Does this challenge the goal?
In-scope gap
It examines adoption barriers and trade-offs for understandability rules within the existing goal of developing practical, risk-aware AI governance. It does not contest that goal.
The question raises possible obstacles; the evidence does not establish that they are decisive or make adoption impossible.
Why this category?
- Consequential implementation and feasibility limits for transparency safeguards
- Trade-offs between understandability and claimed practical benefits
- Possible institutional resistance and secrecy barriers
The evidence is an AI interpretation of one shared source, not independent corroboration. It does not quantify benefits, resistance, or the effectiveness of legal reforms.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
Making and interpreting laws →
Authority, accountability and resilience →
Tests, stop conditions and unresolved work →
Read the exploration goalIdentify how such rules could reduce sought-after benefits or encounter resistance from those who benefit from the systems, without treating adoption as impossible.
What we can say so far
Danaher argues that understandability requirements could reduce some of the practical benefits sought from algorithmic systems, because making machine-learning systems more interpretable can involve trade-offs. Rules may also meet resistance from businesses or governments that benefit from these systems or from keeping them secret—for example, companies may invoke trade secrets and agencies may restrict disclosure to prevent gaming or hacking. These are possible obstacles, not proof that rules cannot be adopted: the paper acknowledges that laws could be changed and that more interpretable designs may be possible, though difficult in practice.
What remains open
- The evidence does not quantify benefit reductions, the prevalence or strength of resistance, or which interests would be decisive.
- It does not assess particular legal reforms or establish that the obstacles cannot be overcome.
Conditions and limits
- These are the author's arguments and reported trade-offs, not measured findings about how often rules face resistance or how much benefit they reduce.
- The evidence is an AI interpretation of the readable author manuscript from an accessible mirror, not independent verification or the publisher's typeset version.
Objections to consider
- The possibility of designing systems that support understanding qualifies any implication that adoption is impossible. The answer preserves that qualification; the evidence also notes that laws could be changed.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
supports
Rules may face resistance Requiring AI decision systems to be understandable may reduce some of their practical benefits and face resistance from businesses or governments that benefit from them.
Directly states that understandability requirements may reduce practical benefits and face resistance from businesses or governments that benefit from the systems.
The Threat of Algocracy: Reality, Resistance and Accommodation
Read full source: The Threat of Algocracy: Reality, Resistance and Accommodation (opens in a new tab) ↗