Australia’s Aged Care Algorithm Is Under Fire. At Last, Someone’s Listening

Why It Matters

Inaccurate algorithmic funding threatens the safety and independence of vulnerable older Australians and exposes the government to legal and reputational risk.

Key Takeaways

• Algorithm determines funding levels for home‑based aged care
• No public evidence validates the tool’s predictive accuracy
• Over 800 seniors have appealed allocations since rollout
• Funding gaps can cost up to $20,000 AUD per year
• Government removed manual overrides, limiting professional judgment

Pulse Analysis

The Support at Home program was introduced in 2025 to keep more seniors living independently, replacing traditional case‑by‑case assessments with a digital Integrated Assessment Tool. The tool aggregates responses from eleven validated questionnaires and feeds them into a proprietary algorithm that assigns one of eight funding tiers, ranging from roughly US$7,000 to US$51,000 annually. While the intent is consistency at scale, the system was deployed without a transparent evidence base linking assessment scores to actual care outcomes, a gap that has drawn scrutiny from policymakers and advocates alike.

Stakeholders point to several technical and ethical flaws. Each of the eleven component tools carries its own error margin, and when combined, these errors compound, undermining the reliability of the final score. Moreover, the algorithm’s black‑box nature prevents families from understanding why a particular funding level was assigned, and the removal of manual override powers leaves clinicians unable to correct obvious mismatches. Financially, a mis‑allocation of just one tier can shift annual support by up to A$20,000 (about US$13,000), a substantial shortfall for those with complex health needs.

The ombudsman’s investigation arrives at a critical juncture, echoing lessons from the Robodebt Royal Commission, which warned that automated decisions must be transparent, fair, and subject to human review. Experts recommend re‑classifying the algorithm as a decision‑support aid rather than a final arbiter, reinstating assessor discretion, and establishing a rigorous validation framework using longitudinal outcome data. Such reforms would restore confidence, protect vulnerable seniors, and align Australia’s aged‑care policy with best practices for algorithmic governance.