Accountability has two faces. Organizations see one. My work reveals the other.

Organizations typically treat accountability as procedure—compliance, documentation, audit. That’s one face, and it’s essential. But accountability also operates through discretionary regimes—the institutional conditions under which professionals have the flexibility to assess, deviate, and decide. When organizations integrate advancing technologies, they risk sustaining the procedural face and eroding the discretionary one. This risk is most acute in critical infrastructure—cybersecurity, financial services, defense, healthcare—where AI-mediated decisions carry consequences that procedural compliance alone cannot manage.

The result looks like governance. Underneath, the discretionary capacity those technologies cannot replace erodes.

Accountability Engineering is the deliberate design of accountability architectures that sustain both regimes.