Westpac, ANZ, NAB appear before parliamentary inquiry into AI — as it happened

TEXT START: Three of the Big Four banks have appeared at a joint parliamentary inquiry into artificial intelligence this afternoon, providing insights into how they are protecting customer data.

The Dissection

The text packages AI deployment as a governance, trust, training, infrastructure, and investment story. Its central fact is more destructive: Westpac reports AI saving 150,000 annualised hours in mortgage and credit assessment, while ANZ cannot provide figures for jobs created. Productivity gains are visible; replacement costs are rhetorically deferred.

Firmus’s failed float does not disprove the Discontinuity Thesis. It shows that speculative capital can misprice AI infrastructure, not that AI loses its structural advantage. A dead IPO is a financing failure, not a resurrection of human labor.

The Core Fallacy

The article treats responsible deployment, worker training, regulation, data sovereignty, and domestic data centres as if they can preserve mass productive participation. They cannot. Under P1–P3, they may slow deployment or relocate ownership, but they do not restore the wage-to-consumption circuit once AI performs cognitive work more cheaply and reliably.

The banks’ denials are especially weak. “No job loss due to AI” is not evidence of no displacement; it is a carefully narrowed statement about attribution. The unanswered question—how many jobs AI actually created—reveals the missing denominator. Hours eliminated are measurable. Replacement employment is not.

Hidden Assumptions

• Training converts displaced workers into economically necessary workers rather than producing a larger queue of credentialed substitutes.
• Regulation can preserve human-only domains at scale despite competitive pressure.
• Domestic data centres distribute AI’s gains broadly instead of concentrating them among capital, energy, land, and infrastructure owners.
• Trust and safety produce legitimacy without addressing ownership of the productive system.
• Offshoring and AI are separable enough that firms can deny the combined labor effect.
• More AI adoption means more jobs, rather than fewer workers supervising more automated output.

Social Function

This is ideological anesthetic and transition management with a partial truth embedded inside it. The partial truth is that banks face real risks around scams, fraud, customer data, cyber-defence, and agentic systems. The anesthetic is the insistence that training, optimism, and “responsible deployment” will reconcile automation with mass employment.

The inquiry also performs elite self-exoneration. Executives present themselves as custodians of trust and opportunity while declining to quantify the labor created by systems already removing labor hours. The language of “skills,” “hope,” and “opportunity” converts a distributional conflict into a public-relations exercise.

The Verdict

The article records the early institutional choreography of displacement, not a solution to it. Banks are building the machinery that erodes human necessity while asking regulators and the public to call the process inclusion, training, and national opportunity. The temporary moats are regulation, liability, customer trust, legacy systems, and scarce infrastructure. They are hospice care for the old employment circuit, not a cure.

The Discontinuity Thesis remains intact: ownership and control move toward Sovereigns; a narrower technical, security, compliance, and infrastructure layer survives as Servitors; the majority are managed as consumers, claimants, or excess capacity. The banks’ inability to show jobs created is the most revealing line in the entire text.