If every automated HR decision still traces back to a human choice, why do we keep saying “the algorithm did it”?
Every module closes on the evidence and the other side. Here that means auditing Murire (2024), adding the Australian legal context the deck omits, and stating the spine plainly: AI in HR is a chain of human decisions wearing a machine's face.
Murire (2024), Administrative Sciences (MDPI) 14:316 — a PRISMA systematic review of how AI reshapes organisational work practices and culture, framed by Schein's three levels of culture (artifacts → espoused values → underlying assumptions). Reported opportunities: efficiency, data-driven decisions, personalised experience, new business models. Reported challenges: cultural alignment, employee resistance, ethical concerns, leadership communication — with transparent leadership and skills investment as the pivotal success strategies.
The set materials are largely US/vendor. Australia is tightening: from 10 December 2026, Privacy Act reforms require organisations to disclose automated decisions that significantly affect people — directly relevant to AI hiring and performance systems. Age discrimination in recruitment is unlawful under the Age Discrimination Act 2004, and the Australian Human Rights Commission has flagged algorithmic hiring bias as a live risk.
The whole module braids into one claim: it's people, not the AI. A human bought the tool, chose the data, set the threshold, and decided whether anyone could appeal. The asymmetry is the injustice — the system sees everything about the applicant; the applicant sees nothing about the system.
The fix is not a better model. It is a named human who is accountable, a candidate who is told a machine was used and can contest it, and a bias audit before deployment — the same governance-first move the back half of the module points to.
Close every HR-AI argument the same way: name the person accountable. “The algorithm decided” is not an answer — it is the problem.
What kind of study is Murire (2024), and what is its main methodological limit?
What changes in Australian law from 10 December 2026, and why does it matter for HR AI?
State the module's accountability spine in one line.