Bias detection and reduction in machine-learning techniques
Summary
The USPTO granted patent US12596957B2 to Equifax, Inc. on April 7, 2026, covering methods for detecting and reducing bias in machine learning models used for risk assessment. The patent describes training ML models using samples, calculating bias metrics from protected attributes, and modifying models based on detected bias to control entity access to interactive computing environments. The patent contains 20 claims and names Mufeng Zou, Swathi Veeravelly, and Marcus Bruhn as inventors.
What changed
The USPTO granted a patent to Equifax, Inc. for a system and method of improving machine learning models for risk assessment by removing or reducing bias. The system trains ML models using samples, obtains protected attribute data, calculates bias metrics, and detects bias based on those metrics. Detected bias triggers model modification and retraining, with the retrained model used to predict risk indicators for target entities. The predicted indicators can be transmitted to remote computing devices for controlling access to interactive computing environments.
Patent grants do not impose compliance obligations on third parties but establish enforceable intellectual property rights for the assignee. Companies developing ML-based risk assessment or bias detection systems should review the patent claims to assess potential licensing needs or design-around considerations. The patent's broad claims covering protected attribute analysis and bias metric calculation in risk assessment contexts may affect product development strategies for competing firms in credit reporting, financial services, or AI technology sectors.
What to do next
- Monitor for updates
Archived snapshot
Apr 7, 2026GovPing captured this document from the original source. If the source has since changed or been removed, this is the text as it existed at that time.
Bias detection and reduction in machine-learning techniques
Grant US12596957B2 Kind: B2 Apr 07, 2026
Assignee
EQUIFAX, INC.
Inventors
Mufeng Zou, Swathi Veeravelly, Marcus Bruhn
Abstract
In some aspects, a computing system can improve a machine learning model for risk assessment by removing or reducing bias in the machine learning model. The training process for the machine learning model can include training the machine learning model using training samples, obtaining data for a protected attribute, and calculating a bias metric using the data for the protected attribute and data obtained from the trained machine learning model. Based on the bias metric, bias associated with the machine learning model can be detected. The machine learning model can be modified based on the detected bias and re-trained. The re-trained machine learning model can be used to predict a risk indicator for a target entity. The predicted risk indicator can be transmitted to a remote computing device and be used for controlling access of the target entity to one or more interactive computing environments.
CPC Classifications
G06N 20/00 G06N 20/10 G06N 5/04 G06N 5/01 G06N 3/098 G06N 3/088 G06N 3/0454-0455 H04L 63/10 G06F 40/44 G06F 18/2155 G06F 18/2113 G06F 18/217 G06Q 10/0635
Filing Date
2022-10-14
Application No.
18046661
Claims
20
Related changes
Get daily alerts for ChangeBridge: Patent Grants - AI & Computing (G06N)
Daily digest delivered to your inbox.
Free. Unsubscribe anytime.
Source
About this page
Every important government, regulator, and court update from around the world. One place. Real-time. Free. Our mission
Source document text, dates, docket IDs, and authority are extracted directly from USPTO.
The plain-English summary, classification, and "what to do next" steps are AI-generated from the original text. Cite the source document, not the AI analysis.
Classification
Who this affects
Taxonomy
Browse Categories
Get alerts for this source
We'll email you when ChangeBridge: Patent Grants - AI & Computing (G06N) publishes new changes.