On Monday, machine-learning solutions provider Stratyfy released UnBias, which helps financial institutions discover and undo bias in their systems. UnBias will be available as a standalone offering.

Why should we care?
If Stratyfy’s provisional results hold as this product scales, then UnBias could contribute to a threefold reduction in bias within financial systems—namely in processes like credit underwriting, which has long been a source of exclusion and discrimination. Now being studied by Stanford and FinRegLab, UnBias may also offer best practices for other tech-driven tools that look to decrease bias within finance. “New regulations and market forces have generated an unprecedented sense of urgency in the need to address bias, but adoption of adequate tools remains a challenge,” Stratyfy said in a press release. Scholars of algorithmic bias tend to advocate for the rollback of AI-driven tools, rather than their improvement. With that in mind, both advocates and critics of financial automation will track UnBias’s progress in the months to come.