Why “human in the loop” is insufficient unless the human has time, information, independence and real authority to change the outcome.
A human must be more than present. Meaningful authority requires relevant evidence, enough time, real override power, independence from the original decision and protection from institutional incentives that make disagreement impossible.
A reviewer handling thousands of cases with seconds per file may formally approve decisions while the machine effectively governs. Concresca therefore measures reviewer authority, not merely reviewer existence.
The EU AI Act includes human-oversight obligations for covered high-risk AI; Canada’s federal directive uses graduated human involvement and recourse; California’s ADMT rules create covered consumer rights. These are current governance facts, not proof that meaningful review is always achieved in practice.
Source links establish traceability and support. They do not imply that the source endorses Concresca’s constitutional proposals.
Includes logging and human-oversight obligations for high-risk AI and prohibitions relevant to certain social-scoring and criminal-risk uses.
European Union — official/source page ↗Requires impact assessment, transparency, quality assurance, recourse, and graduated human involvement for covered federal automated decisions.
Government of Canada — official/source page ↗Documents California ADMT regulations including risk-assessment and consumer-rights requirements.
California Privacy Protection Agency — official/source page ↗NO JUDGMENT WHATSOEVER. Concresca coordinates without assigning moral worth, character, guilt, danger, trustworthiness, loyalty, purity, normality, or social standing. Questions, thoughts, identities, messages, content, and conduct are not objects of Concresca judgment.