United StatesPublic sector2018–2022Oregon Department of Human Services
Safety at Screening child-welfare risk model
Oregon announced the withdrawal of its child-welfare screening algorithm
Problem
Hotline workers needed to decide which reports warranted a child-protection investigation. The agency introduced risk scores to support those decisions.
Approach
The Safety at Screening model predicted future investigations or foster-care placement. Workers saw its numerical scores when deciding whether to investigate. In 2022, officials announced that use would cease by the end of June as Oregon changed its screening process.
Results
No published figures
Takeaway
The agency cited incompatibility with its replacement screening process. Scrutiny of racial disparities preceded the withdrawal, but officials did not disclose Oregon-specific disparities explaining the decision.
Categories
- Size
- Large enterprise
- Industry
- Public sector
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Partner co-development
- Goal
- Quality and risk
- Evidence
- Company's own claim
Sources
Source summaries
The Imprint reported Oregon's announced withdrawal, drawing on Associated Press reporting and agency statements. It described the model's screening role and distinguished concerns about racial bias from the agency's explanation about its replacement process.
Full details are in the original sources.